More Math Into LATEX 4th Edition
George Gr¨atzer
More Math Into LATEX 4th Edition
Foreword by
Rainer Sch¨opf LATEX3 team
George Gr¨atzer Department of Mathematics University of Manitoba Winnipeg, MB R3T 2N2 Canada
[email protected] Cover design by Mary Burgess. Typeset by the author in LATEX. Library of Congress Control Number: 2007923503 ISBN13: 9780387322896
eISBN13: 9780387688527
Printed on acidfree paper. c 2007 Springer Science+Business Media, LLC All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media LLC, 233 Spring Street, New York, NY 10013, USA) and the author, except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
9 8 7 6 5 4 3 2 1 springer.com
(HP)
To the Volunteers without whose dedication over 15 years, this book could not have been done
Danny (11),
and to my four grandchildren
Anna (8), Emma (2), and Kate (0)
Short Contents
Foreword
xxi
Preface to the Fourth Edition
xxv
Introduction
xxix
I Short Course
1
1
Your LATEX
3
2
Typing text
7
3
Typing math
17
4
Your first article and presentation
35
II Text and Math
59
5
Typing text
61
6
Text environments
117
7
Typing math
151
8
More math
187
9
Multiline math displays
207
viii
Short Contents
III
Document Structure
245
10 LATEX documents
247
11 The AMS article document class
271
12 Legacy document classes
303
IV
315
Presentations and PDF Documents
13 PDF documents
317
14 Presentations
325
V
Customization
361
15 Customizing LATEX
363
VI
419
Long Documents
16 B IBTEX
421
17 MakeIndex
449
18 Books in LATEX
465
A Installation
489
B Math symbol tables
501
C Text symbol tables
515
D Some background
521
E LATEX and the Internet
537
F PostScript fonts
543
G LATEX localized
547
H Final thoughts
551
Bibliography
557
Index
561
Contents Foreword
xxi
Preface to the Fourth Edition xxv Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii Introduction xxix Is this book for you? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxix
I Short Course
1
1
. . . .
3 3 4 4 5
2
3
Your LATEX 1.1 Your computer . . . . 1.2 Sample files . . . . . . 1.3 Editing cycle . . . . . 1.4 Three productivity tools Typing text 2.1 The keyboard . . 2.2 Your first note . . 2.3 Lines too wide . . 2.4 More text features
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
7 8 9 12 13
Typing math 3.1 A note with math . . . . . . 3.2 Errors in math . . . . . . . . 3.3 Building blocks of a formula 3.4 Displayed formulas . . . . . 3.4.1 Equations . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
17 17 19 22 27 27
. . . .
. . . .
. . . .
. . . .
. . . .
x
Contents 3.4.2 3.4.3 4
Aligned formulas . . . . . . . . . . . . . . . . . . . . . . . . Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Your first article and presentation 4.1 The anatomy of an article . . . . . . . 4.1.1 The typeset sample article . . 4.2 An article template . . . . . . . . . . 4.2.1 Editing the top matter . . . . 4.2.2 Sectioning . . . . . . . . . . 4.2.3 Invoking proclamations . . . . 4.2.4 Inserting references . . . . . . 4.3 On using LATEX . . . . . . . . . . . . 4.3.1 LATEX error messages . . . . . 4.3.2 Logical and visual design . . . 4.4 Converting an article to a presentation 4.4.1 Preliminary changes . . . . . 4.4.2 Making the pages . . . . . . . 4.4.3 Fine tuning . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
II Text and Math 5
Typing text 5.1 The keyboard . . . . . . . . . . . . . 5.1.1 Basic keys . . . . . . . . . . 5.1.2 Special keys . . . . . . . . . 5.1.3 Prohibited keys . . . . . . . . 5.2 Words, sentences, and paragraphs . . 5.2.1 Spacing rules . . . . . . . . . 5.2.2 Periods . . . . . . . . . . . . 5.3 Commanding LATEX . . . . . . . . . . 5.3.1 Commands and environments 5.3.2 Scope . . . . . . . . . . . . . 5.3.3 Types of commands . . . . . 5.4 Symbols not on the keyboard . . . . . 5.4.1 Quotation marks . . . . . . . 5.4.2 Dashes . . . . . . . . . . . . 5.4.3 Ties or nonbreakable spaces . 5.4.4 Special characters . . . . . . . 5.4.5 Ellipses . . . . . . . . . . . . 5.4.6 Ligatures . . . . . . . . . . . 5.4.7 Accents and symbols in text . 5.4.8 Logos and dates . . . . . . . .
30 33 35 35 41 44 44 46 46 47 48 48 52 53 53 55 55
59 . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
61 62 62 63 63 64 64 66 67 68 71 73 74 75 75 76 76 78 79 79 80
xi
Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. 82 . 85 . 85 . 87 . 88 . 88 . 89 . 90 . 91 . 93 . 93 . 94 . 94 . 95 . 95 . 96 . 99 . 100 . 101 . 102 . 102 . 104 . 105 . 106 . 107 . 107 . 109 . 110 . 112 . 113 . 115
Text environments 6.1 Some general rules for displayed text environments 6.2 List environments . . . . . . . . . . . . . . . . . . 6.2.1 Numbered lists . . . . . . . . . . . . . . . 6.2.2 Bulleted lists . . . . . . . . . . . . . . . . 6.2.3 Captioned lists . . . . . . . . . . . . . . . 6.2.4 A rule and combinations . . . . . . . . . . 6.3 Style and size environments . . . . . . . . . . . . 6.4 Proclamations (theoremlike structures) . . . . . . 6.4.1 The full syntax . . . . . . . . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
5.5
5.6
5.7
5.8
5.9
6
5.4.9 Hyphenation . . . . . . . . . Comments and footnotes . . . . . . . 5.5.1 Comments . . . . . . . . . . 5.5.2 Footnotes . . . . . . . . . . . Changing font characteristics . . . . . 5.6.1 Basic font characteristics . . . 5.6.2 Document font families . . . . 5.6.3 Shape commands . . . . . . . 5.6.4 Italic corrections . . . . . . . 5.6.5 Series . . . . . . . . . . . . . 5.6.6 Size changes . . . . . . . . . 5.6.7 Orthogonality . . . . . . . . . 5.6.8 Obsolete twoletter commands 5.6.9 Lowlevel commands . . . . . Lines, paragraphs, and pages . . . . . 5.7.1 Lines . . . . . . . . . . . . . 5.7.2 Paragraphs . . . . . . . . . . 5.7.3 Pages . . . . . . . . . . . . . 5.7.4 Multicolumn printing . . . . . Spaces . . . . . . . . . . . . . . . . . 5.8.1 Horizontal spaces . . . . . . . 5.8.2 Vertical spaces . . . . . . . . 5.8.3 Relative spaces . . . . . . . . 5.8.4 Expanding spaces . . . . . . . Boxes . . . . . . . . . . . . . . . . . 5.9.1 Line boxes . . . . . . . . . . 5.9.2 Frame boxes . . . . . . . . . 5.9.3 Paragraph boxes . . . . . . . 5.9.4 Marginal comments . . . . . 5.9.5 Solid boxes . . . . . . . . . . 5.9.6 Fine tuning boxes . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
117 118 118 119 119 120 120 123 124 128
xii
Contents 6.4.2 Proclamations with style . . . . . . Proof environments . . . . . . . . . . . . . Tabular environments . . . . . . . . . . . . 6.6.1 Table styles . . . . . . . . . . . . . Tabbing environments . . . . . . . . . . . . Miscellaneous displayed text environments
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
129 131 133 140 141 143
Typing math 7.1 Math environments . . . . . . . . . . . . . 7.2 Spacing rules . . . . . . . . . . . . . . . . 7.3 Equations . . . . . . . . . . . . . . . . . . 7.4 Basic constructs . . . . . . . . . . . . . . . 7.4.1 Arithmetic operations . . . . . . . . 7.4.2 Binomial coefficients . . . . . . . . 7.4.3 Ellipses . . . . . . . . . . . . . . . 7.4.4 Integrals . . . . . . . . . . . . . . . 7.4.5 Roots . . . . . . . . . . . . . . . . 7.4.6 Text in math . . . . . . . . . . . . 7.4.7 Building a formula stepbystep . . 7.5 Delimiters . . . . . . . . . . . . . . . . . . 7.5.1 Stretching delimiters . . . . . . . . 7.5.2 Delimiters that do not stretch . . . . 7.5.3 Limitations of stretching . . . . . . 7.5.4 Delimiters as binary relations . . . 7.6 Operators . . . . . . . . . . . . . . . . . . 7.6.1 Operator tables . . . . . . . . . . . 7.6.2 Defining operators . . . . . . . . . 7.6.3 Congruences . . . . . . . . . . . . 7.6.4 Large operators . . . . . . . . . . . 7.6.5 Multiline subscripts and superscripts 7.7 Math accents . . . . . . . . . . . . . . . . 7.8 Stretchable horizontal lines . . . . . . . . . 7.8.1 Horizontal braces . . . . . . . . . . 7.8.2 Overlines and underlines . . . . . . 7.8.3 Stretchable arrow math symbols . . 7.9 Formula Gallery . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .
151 152 154 156 157 157 159 160 161 161 162 164 166 167 168 169 170 170 171 173 173 174 176 176 178 178 179 179 180
. . . . .
187 187 188 188 190 190
6.5 6.6 6.7 6.8 7
8
More math 8.1 Spacing of symbols . . . . 8.1.1 Classification . . . 8.1.2 Three exceptions . 8.1.3 Spacing commands 8.1.4 Examples . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
xiii
Contents 8.1.5 The phantom command . . . . . Building new symbols . . . . . . . . . . 8.2.1 Stacking symbols . . . . . . . . . 8.2.2 Negating and sidesetting symbols 8.2.3 Changing the type of a symbol . . Math alphabets and symbols . . . . . . . 8.3.1 Math alphabets . . . . . . . . . . 8.3.2 Math symbol alphabets . . . . . . 8.3.3 Bold math symbols . . . . . . . . 8.3.4 Size changes . . . . . . . . . . . 8.3.5 Continued fractions . . . . . . . . Vertical spacing . . . . . . . . . . . . . . Tagging and grouping . . . . . . . . . . . Miscellaneous . . . . . . . . . . . . . . . 8.6.1 Generalized fractions . . . . . . . 8.6.2 Boxed formulas . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . .
191 192 192 194 195 195 196 197 197 199 200 200 201 204 204 205
Multiline math displays 9.1 Visual Guide . . . . . . . . . . . . . . . 9.1.1 Columns . . . . . . . . . . . . . 9.1.2 Subsidiary math environments . . 9.1.3 Adjusted columns . . . . . . . . 9.1.4 Aligned columns . . . . . . . . . 9.1.5 Touring the Visual Guide . . . . . 9.2 Gathering formulas . . . . . . . . . . . . 9.3 Splitting long formulas . . . . . . . . . . 9.4 Some general rules . . . . . . . . . . . . 9.4.1 General rules . . . . . . . . . . . 9.4.2 Subformula rules . . . . . . . . . 9.4.3 Breaking and aligning formulas . 9.4.4 Numbering groups of formulas . . 9.5 Aligned columns . . . . . . . . . . . . . 9.5.1 An align variant . . . . . . . . . 9.5.2 eqnarray, the ancestor of align 9.5.3 The subformula rule revisited . . 9.5.4 The alignat environment . . . . 9.5.5 Inserting text . . . . . . . . . . . 9.6 Aligned subsidiary math environments . . 9.6.1 Subsidiary variants . . . . . . . . 9.6.2 Split . . . . . . . . . . . . . . . . 9.7 Adjusted columns . . . . . . . . . . . . . 9.7.1 Matrices . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . .
207 207 209 209 210 210 210 211 212 215 215 215 217 218 219 221 222 223 224 226 227 227 230 231 232
8.2
8.3
8.4 8.5 8.6
9
xiv
Contents
9.8 9.9
III
9.7.2 Arrays . . . . . 9.7.3 Cases . . . . . Commutative diagrams Adjusting the display .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. 236 . 239 . 240 . 242
Document Structure
10 LATEX documents 10.1 The structure of a document . . . . . . 10.2 The preamble . . . . . . . . . . . . . . 10.3 Top matter . . . . . . . . . . . . . . . . 10.3.1 Abstract . . . . . . . . . . . . . 10.4 Main matter . . . . . . . . . . . . . . . 10.4.1 Sectioning . . . . . . . . . . . 10.4.2 Crossreferencing . . . . . . . . 10.4.3 Floating tables and illustrations 10.5 Back matter . . . . . . . . . . . . . . . 10.5.1 Bibliographies in articles . . . . 10.5.2 Simple indexes . . . . . . . . . 10.6 Visual design . . . . . . . . . . . . . .
245 . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . .
11 The AMS article document class 11.1 Why amsart? . . . . . . . . . . . . . . . . . . . . 11.1.1 Submitting an article to the AMS . . . . . 11.1.2 Submitting an article to Algebra Universalis 11.1.3 Submitting to other journals . . . . . . . . 11.1.4 Submitting to conference proceedings . . . 11.2 The top matter . . . . . . . . . . . . . . . . . . . . 11.2.1 Article information . . . . . . . . . . . . . 11.2.2 Author information . . . . . . . . . . . . . 11.2.3 AMS information . . . . . . . . . . . . . . 11.2.4 Multiple authors . . . . . . . . . . . . . . 11.2.5 Examples . . . . . . . . . . . . . . . . . . 11.2.6 Abstract . . . . . . . . . . . . . . . . . . . 11.3 The sample article . . . . . . . . . . . . . . . . . . 11.4 Article templates . . . . . . . . . . . . . . . . . . 11.5 Options . . . . . . . . . . . . . . . . . . . . . . . 11.6 The AMS packages . . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . . . . . . . . . . .
. . . . . . . . . . . .
247 248 249 251 251 251 252 255 258 261 261 267 268
. . . . . . . . . . . . . . . .
271 271 271 272 272 273 273 273 275 279 281 282 285 285 294 297 300
xv
Contents 12 Legacy document classes 12.1 Articles and reports . . 12.1.1 Top matter . . 12.1.2 Options . . . . 12.2 Letters . . . . . . . . . 12.3 The LATEX distribution . 12.3.1 Tools . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
303 . 303 . 304 . 306 . 308 . 310 . 312
IV Presentations and PDF Documents 13 PDF documents 13.1 PostScript and PDF . . . . . . . . 13.1.1 PostScript . . . . . . . . . 13.1.2 PDF . . . . . . . . . . . . 13.1.3 Hyperlinks . . . . . . . . 13.2 Hyperlinks for LATEX . . . . . . . 13.2.1 Using hyperref . . . . . 13.2.2 backref and colorlinks 13.2.3 Bookmarks . . . . . . . . 13.2.4 Additional commands . .
315 . . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
317 317 317 318 319 319 320 320 321 322
14 Presentations 14.1 Quick and dirty beamer . . . . . . . . . . . . . . . . . 14.1.1 First changes . . . . . . . . . . . . . . . . . . 14.1.2 Changes in the body . . . . . . . . . . . . . . 14.1.3 Making things prettier . . . . . . . . . . . . . 14.1.4 Adjusting the navigation . . . . . . . . . . . . 14.2 Baby beamers . . . . . . . . . . . . . . . . . . . . . 14.2.1 Overlays . . . . . . . . . . . . . . . . . . . . 14.2.2 Understanding overlays . . . . . . . . . . . . . 14.2.3 More on the \only and \onslide commands . 14.2.4 Lists as overlays . . . . . . . . . . . . . . . . 14.2.5 Out of sequence overlays . . . . . . . . . . . . 14.2.6 Blocks and overlays . . . . . . . . . . . . . . 14.2.7 Links . . . . . . . . . . . . . . . . . . . . . . 14.2.8 Columns . . . . . . . . . . . . . . . . . . . . 14.2.9 Coloring . . . . . . . . . . . . . . . . . . . . 14.3 The structure of a presentation . . . . . . . . . . . . . 14.3.1 Longer presentations . . . . . . . . . . . . . . 14.3.2 Navigation symbols . . . . . . . . . . . . . . 14.4 Notes . . . . . . . . . . . . . . . . . . . . . . . . . . 14.5 Themes . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . .
325 326 326 327 328 328 333 333 335 337 339 341 343 343 347 348 350 354 354 355 356
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
. . . . . . . . .
xvi
Contents 14.6 Planning your presentation . . . . . . . . . . . . . . . . . . . . . . . 14.7 What did I leave out? . . . . . . . . . . . . . . . . . . . . . . . . . .
V
Customization
361
15 Customizing LATEX 15.1 Userdefined commands . . . . . . . . . . . . . . . 15.1.1 Examples and rules . . . . . . . . . . . . . . 15.1.2 Arguments . . . . . . . . . . . . . . . . . . 15.1.3 Short arguments . . . . . . . . . . . . . . . 15.1.4 Optional arguments . . . . . . . . . . . . . . 15.1.5 Redefining commands . . . . . . . . . . . . 15.1.6 Redefining names . . . . . . . . . . . . . . . 15.1.7 Showing the definitions of commands . . . . 15.1.8 Delimited commands . . . . . . . . . . . . . 15.2 Userdefined environments . . . . . . . . . . . . . . 15.2.1 Modifying existing environments . . . . . . 15.2.2 Arguments . . . . . . . . . . . . . . . . . . 15.2.3 Optional arguments with default values . . . 15.2.4 Short contents . . . . . . . . . . . . . . . . 15.2.5 Brandnew environments . . . . . . . . . . . 15.3 A custom command file . . . . . . . . . . . . . . . . 15.4 The sample article with userdefined commands . . . 15.5 Numbering and measuring . . . . . . . . . . . . . . 15.5.1 Counters . . . . . . . . . . . . . . . . . . . 15.5.2 Length commands . . . . . . . . . . . . . . 15.6 Custom lists . . . . . . . . . . . . . . . . . . . . . . 15.6.1 Length commands for the list environment 15.6.2 The list environment . . . . . . . . . . . . 15.6.3 Two complete examples . . . . . . . . . . . 15.6.4 The trivlist environment . . . . . . . . . 15.7 The dangers of customization . . . . . . . . . . . . .
VI
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . .
Long Documents
16 B IBTEX 16.1 The database . . . . 16.1.1 Entry types . 16.1.2 Typing fields 16.1.3 Articles . . . 16.1.4 Books . . . .
358 358
. . . . . . . . . . . . . . . . . . . . . . . . . .
363 364 364 370 373 374 374 375 376 378 380 380 383 384 385 385 386 392 398 399 403 406 407 409 411 414 415
419 . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
421 . 423 . 423 . 426 . 428 . 429
xvii
Contents 16.1.5 Conference proceedings and collections 16.1.6 Theses . . . . . . . . . . . . . . . . . 16.1.7 Technical reports . . . . . . . . . . . . 16.1.8 Manuscripts and other entry types . . . 16.1.9 Abbreviations . . . . . . . . . . . . . . 16.2 Using B IBTEX . . . . . . . . . . . . . . . . . . 16.2.1 Sample files . . . . . . . . . . . . . . . 16.2.2 Setup . . . . . . . . . . . . . . . . . . 16.2.3 Four steps of B IBTEXing . . . . . . . . 16.2.4 B IBTEX rules and messages . . . . . . 16.2.5 Submitting an article . . . . . . . . . . 16.3 Concluding comments . . . . . . . . . . . . . 17 MakeIndex 17.1 Preparing the document . . . 17.2 Index commands . . . . . . 17.3 Processing the index entries . 17.4 Rules . . . . . . . . . . . . 17.5 Multiple indexes . . . . . . 17.6 Glossary . . . . . . . . . . . 17.7 Concluding comments . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . .
. . . . . . .
18 Books in LATEX 18.1 Book document classes . . . . . . . . . . . . . . . 18.1.1 Sectioning . . . . . . . . . . . . . . . . . 18.1.2 Division of the body . . . . . . . . . . . . 18.1.3 Document class options . . . . . . . . . . 18.1.4 Title pages . . . . . . . . . . . . . . . . . 18.1.5 Springer’s document class for monographs 18.2 Tables of contents, lists of tables and figures . . . . 18.2.1 Tables of contents . . . . . . . . . . . . . 18.2.2 Lists of tables and figures . . . . . . . . . 18.2.3 Exercises . . . . . . . . . . . . . . . . . . 18.3 Organizing the files for a book . . . . . . . . . . . 18.3.1 The folders and the master document . . . 18.3.2 Inclusion and selective inclusion . . . . . . 18.3.3 Organizing your files . . . . . . . . . . . . 18.4 Logical design . . . . . . . . . . . . . . . . . . . . 18.5 Final preparations for the publisher . . . . . . . . . 18.6 If you create the PDF file for your book . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . .
. 430 . 433 . 434 . 435 . 436 . 437 . 437 . 439 . 440 . 443 . 446 . 446
. . . . . . .
. . . . . . .
449 449 453 459 462 463 464 464
. . . . . . . . . . . . . . . . .
465 466 466 467 468 469 469 473 473 475 476 476 477 478 479 479 482 484
. . . . . . . . . . . . . . . . .
xviii
Contents A Installation A.1 LATEX on a PC . . . . . . . . . . A.1.1 Installing MiKTeX . . . . A.1.2 Installing WinEdt . . . . A.1.3 The editing cycle . . . . A.1.4 Making a mistake . . . . A.1.5 Three productivity tools A.1.6 An important folder . . . A.2 LATEX on a Mac . . . . . . . . . A.2.1 Installations . . . . . . . A.2.2 Working with TeXShop . A.2.3 The editing cycle . . . . A.2.4 Making a mistake . . . . A.2.5 Three productivity tools A.2.6 An important folder . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
. . . . . . . . . . . . . .
489 490 490 490 491 491 494 494 495 495 496 498 498 498 499
B Math symbol tables B.1 Hebrew and Greek letters B.2 Binary relations . . . . . B.3 Binary operations . . . . B.4 Arrows . . . . . . . . . B.5 Miscellaneous symbols . B.6 Delimiters . . . . . . . . B.7 Operators . . . . . . . . B.7.1 Large operators . B.8 Math accents and fonts . B.9 Math spacing commands
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
501 501 503 506 507 508 509 510 511 512 513
C Text symbol tables C.1 Some European characters . . . . . . . . C.2 Text accents . . . . . . . . . . . . . . . . C.3 Text font commands . . . . . . . . . . . . C.3.1 Text font family commands . . . C.3.2 Text font size changes . . . . . . C.4 Additional text symbols . . . . . . . . . . C.5 Additional text symbols with T1 encoding C.6 Text spacing commands . . . . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
515 515 516 516 516 517 518 519 520
D Some background D.1 A short history . . . . . . . . . . D.1.1 TEX . . . . . . . . . . . . D.1.2 LATEX 2.09 and AMSTEX D.1.3 LATEX3 . . . . . . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
521 . 521 . 521 . 522 . 523
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . .
. . . .
. . . .
. . . .
xix
Contents D.1.4 More recent developments D.2 Structure . . . . . . . . . . . . . . D.2.1 Using LATEX . . . . . . . . D.2.2 AMS packages revisited . D.3 How LATEX works . . . . . . . . . D.3.1 The layers . . . . . . . . . D.3.2 Typesetting . . . . . . . . D.3.3 Viewing and printing . . . D.3.4 LATEX’s files . . . . . . . . D.4 Interactive LATEX . . . . . . . . . D.5 Separating form and content . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
. . . . . . . . . . .
524 525 525 528 528 528 529 530 531 534 535
E LATEX and the Internet 537 E.1 Obtaining files from the Internet . . . . . . . . . . . . . . . . . . . . 537 E.2 The TEX Users Group . . . . . . . . . . . . . . . . . . . . . . . . . . 541 E.3 Some useful sources of LATEX information . . . . . . . . . . . . . . . 542 F PostScript fonts 543 F.1 The Times font and MathTıme . . . . . . . . . . . . . . . . . . . . . 544 F.2 Lucida Bright fonts . . . . . . . . . . . . . . . . . . . . . . . . . . . 546 F.3 More PostScript fonts . . . . . . . . . . . . . . . . . . . . . . . . . . 546 G LATEX localized
547
H Final thoughts H.1 What was left out? . . . H.1.1 LATEX omissions H.1.2 TEX omissions . H.2 Further reading . . . . . H.3 What’s coming . . . . .
551 551 551 552 553 554
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
Bibliography
557
Index
561
Foreword It was the autumn of 1989—a few weeks before the Berlin wall came down, President George H. W. Bush was president, and the American Mathematical Society decided to outsource TEX programming to Frank Mittelbach and me. Why did the AMS outsource TEX programming to us? This was, after all, a decade before the words “outsourcing” and “offshore” entered the lexicon. There were many American TEX experts. Why turn elsewhere? For a number of years, the AMS tried to port the mathematical typesetting features of AMSTEX to LATEX, but they made little progress with the AMSFonts. Frank and I had just published the New Font Selection Scheme for LATEX, which went a long way to satisfy what they wanted to accomplish. So it was logical that the AMS turned to us to add AMSFonts to LATEX. Being young and enthusiastic, we convinced the AMS that the AMSTEX commands should be changed to conform to the LATEX standards. Michael Downes was assigned as our AMS contact; his insight was a tremendous help. We already had LATEXNFSS, which could be run in two modes: compatible with the old LATEX or enabled with the new font features. We added the reworked AMSTEX code to LATEXNFSS, thus giving birth to AMSLATEX, released by the AMS at the August 1990 meeting of the International Mathematical Union in Kyoto. AMSLATEX was another variant of LATEX. Many installations had several LATEX variants to satisfy the needs of their users: with old and new font changing commands, with and without AMSLATEX, a single and a multilanguage version. We decided to develop a Standard LATEX that would reconcile all the variants. Out of a group of interested people grew what was later called the LATEX3 team—and the LATEX3 project got underway. The team’s first major accomplishment was the release of LATEX 2ε in June 1994. This standard LATEX incorporates all the improvements we wanted back in 1989. It is now very stable and it is uniformly used. Under the direction of Michael Downes, our AMSLATEX code was turned into AMS packages that run under LATEX just like other packages. Of course, the LATEX3
xxii
Foreword team recognizes that these are special; we call them “required packages” because they are part and parcel of a mathematician’s standard toolbox. Since then a lot has been achieved to make an author’s task easier. A tremendous number of additional packages are available today. The LATEX Companion, 2nd edition, describes many of my favorite packages. George Gr¨atzer got involved with these developments in 1990, when he got his copy of AMSLATEX in Kyoto. The documentation he received explained that AMSLATEX is a LATEX variant—read Lamport’s LATEX book to get the proper background. AMSLATEX is not AMSTEX either—read Spivak’s AMSTEX book to get the proper background. The rest of the document explained in what way AMSLATEX differs from LATEX and AMSTEX. Talk about a steep learning curve . . . Luckily, George’s frustration working through this nightmare was eased by a lengthy email correspondence with Frank and lots of telephone calls to Michael. Three years of labor turned into his first book on LATEX, providing a “simple introduction to AMSLATEX”. This fourth edition is more mature, but preserves what made his first book such a success. Just as in the first book, Part I is a short introduction for the beginner, dramatically reducing the steep learning curve of a few weeks to a few hours. The rest of the book is a detailed presentation of what you may need to know. George “teaches by example”. You find in this book many illustrations of even the simplest concepts. For articles, he presents the LATEX source file and the typeset result sidebyside. For formulas, he discusses the building blocks with examples, presents a Formula Gallery, and a Visual Guide to multiline formulas. Going forth and creating “masterpieces of the typesetting art”—as Donald Knuth put it at the end of the TEXbook—requires a fair bit of initiation. This is the book for the LATEX beginner as well as for the advanced user. You just start at a different point. The topics covered include everything you need for mathematical publishing. Starting from scratch, by installing and running LATEX on your own computer Instructions on creating articles, from the simple to the complex Converting an article to a presentation Customize LATEX to your own needs The secrets of writing a book Where to turn to get more information or to download updates The many examples are complemented by a number of easily recognizable features: Rules which you must follow Tips on how to achieve some specific results Experiments to show what happens when you make mistakes—sometimes, it can be difficult to understand what went wrong when all you see is an obscure LATEX error message
xxiii
Foreword
This book teaches you how to convert your mathematical masterpieces into typographical ones, giving you a lot of useful advice on the way. How to avoid the traps for the unwary and how to make your editor happy. And hopefully, you’ll experience the fascination of doing it right. Using good typography to better express your ideas. If you want to learn LATEX, buy this book and start with the Short Course. If you can have only one book on LATEX next to your computer, this is the one to have. And if you want to learn about the world of LATEX packages, also buy a second book, the LATEX Companion, 2nd edition.
Rainer Sch¨opf LATEX3 team
Preface to the Fourth Edition This is my fourth fullsized book on LATEX. The first book, Math into TEX: A Simple Introduction to AMSLATEX [19], written in 1991 and 1992, introduced the brand new AMSLATEX, a LATEX variant not compatible with the LATEX of the time, LATEX 2.09. It brought together the features of LATEX and the math typesetting abilities of AMSTEX, the AMS typesetting language. The second book, Math into LATEX: An Introduction to LATEX and AMSLATEX [27], written in 1995, describes the new LATEX introduced by the LATEX3 team and the AMS typesetting features implemented as extensions of LATEX, called packages. The third book, Math into LATEX, 3rd edition [30], published in 2000, reports on the same system. By 2000, both the “new” LATEX and the AMS packages were quite mature. The feverish debugging of the new LATEX every six months bore fruit. LATEX became very stable. It has changed little since 2000. Version 2.0 of the AMS packages was released and it also became very stable. The third book reports on a rock solid typesetting system. What also changed between 1995 and 2000 is the widespread use of the Internet. Several chapters of the third book deal with the impact of the Internet on mathematical publications. Now, seven years later, we can still report that LATEX—no longer new—and the AMS packages have changed very little. However, the impact of the Internet became even more important. Computers also changed. They are now much more powerful. When I started typesetting math with LATEX, it took two and a half minutes to typeset a page. This book takes 1.8 seconds to typeset on my computer, a Mac desktop from 2006. As a result, we do not have to be very selective in what we load into memory; we can load everything we may possibly need.
xxvi
Preface Circumincession So this is the first big change compared to the previous books. In this book, we roll TEX, LATEX, and the AMS packages into one, and we call it simply LATEX. This results in a great simplification in the exposition and makes the learning curve a little less steep. I am sure with some advanced users this will prove to be a controversial decision. They want to know where a command is defined. For the beginner and the nonexpert user this does not make any difference. What matters is that the command they need be available when they need it. From the beginner’s point of view, this approach is very beneficial. Take as an example the \text command. In all three of my books, we first introduce the LATEX command \mbox for typing text in math formulas. After half a page of discussion comes the sentence: “It is better to enter text in formulas with the \text command provided by the amsmath package.” Then another half page discusses the command \text. In this book, we ignore \mbox and go rightaway to \text. You do not have to do anything to access the command, the amsmath package is always loaded for you. And what to do if you want to find out where a command is defined. Now for both the PC and the Mac, you can easily search for contents of files. Do you want to know where a command is defined? Search for it and it is easy to find the file in which it is introduced. Presentations The second big change is the widespread acceptance of the Adobe PDF format. As a result, the majority of the lectures today at math meetings are given as presentations, PDF files projected to screens using computers. Blackboards and whiteboards have largely disappeared and computer projections are overtaking projectors. So this book takes up presentations as a major topic, introducing it in Part I and discussing it in detail in Chapter 14. Installations In the third book, I report a recurring question that comes up from my readers again and again: Can you help me get started from scratch, covering everything from installing a working LATEX system to the rudiments of text editing? And here is the third big change that has happened in the last few years. While earlier there were dozens of different LATEX implementations and hundreds of text editors, today most PC users use MiKTeX with the text editor/front end WinEdt and most Mac users use TEX Live with the text editor/front end TeXShop. So if you want help to
xxvii
Preface
install LATEX, it is easy for me to help you. Appendix A provides instructions on how to install these systems.
Acknowledgments This book is based, of course, on the three previous books. I would like to thank the many people who read and reread those earlier manuscripts. The editors Richard Ribstein, Thomas R. Scavo, Claire M. Connelly. The professionals Michael Downes (the project leader for the AMS), Frank Mittelbach and David Carlisle (of the LATEX3 team) read and criticized some or all of the three books. Oren Patashnik (the author of B IBTEX) carefully corrected the B IBTEX chapter for two editions. Sebastian Rahtz (the author of the hyperref package and coauthor of The LATEX Web Companion [18]) read the chapter on the Web in the third book. Last but not least, Barbara Beeton of the AMS read all three books with incredible insight. The volunteers for the second book alone, there were 29—listed there. The volunteer readers made tremendous contributions and offered hundreds of pages of corrections. No expert can substitute for the diverse points of view I got from them. My colleagues especially Michael Doob, Harry Lakser, and Craig Platt, who have been very generous with their time. The publishers Edwin Beschler, who believed in the project from the very beginning and guided it through a decade and Ann Kostant who continued Edwin’s work. For this book, I have had the most talented and thorough group of readers ever: Andrew Adler of the University of British Columbia, Canada, Joseph Maria Font of the University of Barcelona, Spain, and Alan Litchfield, of the Auckland University of Technology, New Zealand. Chapter 14 was read by David Derbes, Adam Goldstein, Mark Eli Kalderon, Michael Kubovy, Matthieu Masquelet, and Charilaos Skiadas— and Chapter 15 by Ross Moore. Interestingly, only half of them are mathematicians, the rest are philosophers, linguists, and so on. Appendix A.1 was read by Brian Davey and Appendix A.2 by Richard Koch (the author of TeXShop). The fourth edition was edited by Barbara Beeton, Edwin Beschler, and Clay Martin with Ann Kostant as the Springer editor. The roles of Edwin and Ann have changed, but not the importance of their contributions. The index was compiled with painstaking precision by Laura Kirkland. Barbara Beeton also provided a number of intriguing illustrations of quaint commands. My indebtedness to her cannot be overstated. George Gr¨atzer
Introduction Is this book for you? This book is for the mathematician, physicist, engineer, scientist, linguist, or technical typist who has to learn how to typeset articles containing mathematical formulas or diacritical marks. It teaches you how to use LATEX, a typesetting markup language based on Donald E. Knuth’s typesetting language TEX, designed and implemented by Leslie Lamport, and greatly improved by the AMS. Part I provides a quick introduction to LATEX, from typing examples of text and math to typing your first article (such as the sample article on pages 42–43) and creating your first presentation (such as the sample presentation on pages 57–58) in a very short time. The rest of the book provides a detailed exposition of LATEX. LATEX has a huge collection of rules and commands. While the basics in Part I should serve you well in all your writings, most articles and presentations also require you to look up special topics. Learn Part I well and become passingly familiar enough with the rest of the book, so when the need arises you know where to turn with your problems. You can find specific topics in one or more of the following sources: the Short Contents, the detailed Contents, and the Index.
What is document markup? When you work with a word processor, you see your document on the computer monitor more or less as it looks when printed, with its various fonts, font sizes, font shapes (e.g., roman, italic) and weights (e.g., normal, boldface), interline spacing, indentation, and so on. Working with a markup language is different. You type the source file of your article in a text editor, in which all characters appear in the same font. To indicate changes in the typeset text, you must add text markup commands to the source file.
xxx
Introduction For instance, to emphasize the phrase detailed description in a LATEX source file, type \emph{detailed description} The \emph command is a markup command. The markedup text yields the typeset output detailed description In order to typeset math, √ need math markup commands. As a simple example, R you you may need the formula α2 + x2 dx in an article you are writing. To mark up this formula in LATEX, type $\int \sqrt{\alpha^{2} + x^{2}}\,dx$ You do not have to worry about determining the size of the integral symbol or how to construct the square root symbol that covers α2 + x2 . LATEX does it all for you. On pages 290–293, I juxtapose the source file for a sample article with the typeset version. The markup in the source file may appear somewhat challenging at first, but I think you agree that the typeset article is a pleasing rendering of the original input.
The three layers The markup language we shall discuss comes in three layers: TEX, LATEX, and the AMS packages, described in detail in Appendix D. Most LATEX installations—including the two covered in Appendix A—automatically place all three on your computer. You do not have to know what comes from which layer, so we consider the three together and call it LATEX.
The three platforms Most of you run LATEX on one of the following three computer types: A PC, a computer running Microsoft Windows A Mac1 , a Macintosh computer running OS X A computer running a UNIX variant such as Solaris or Linux The LATEX source file and the typeset version both look the same independent of what computer you have. However, the way you type your source file, the way you typeset it, and the way you look at the typeset version depends on the computer and on the LATEX implementation you use. In Appendix A, we show you how to install LATEX for a PC and a Mac. Many UNIX systems come with LATEX installed. 1 In the old days, I used to run T EXTURES under OS 9. Unfortunately, T EXTURES does not run on new Intel Macs.
Introduction
xxxi
What’s in the book? Part I is the Short Course; it helps you to get started quickly with LATEX, to type your first articles, to prepare your first presentations, and it prepares you to tackle LATEX in more depth in the subsequent parts. We assume here that LATEX is installed on your computer. If it is not, jump to Appendix A. Chapter 1 introduces the terminology we need to talk about your LATEX implementations. Chapter 2 introduces how LATEX uses the keyboard and how to type text. You do not need to learn much to understand the basics. Text markup is quite easy. You learn math markup—which is not so straightforward—in Chapter 3. Several sections in this chapter ease you into mathematical typesetting. There is a section on the basic building blocks of math formulas. Another one discusses equations. Finally, we present the two simplest multiline formulas, which, however, cover most of your everyday needs. In Chapter 4, you start writing your first article and prepare your first presentation. A LATEX article is introduced with the sample article intrart.tex. We analyze in detail its structure and its source file, and we look at the typeset version. Based on this, we prepare an article template, and you are ready for your first article. A quick conversion of the article intrart.tex to a presentation introduces this important topic. Part II introduces the two most basic skills for writing with LATEX in depth, typing text and typing math. Chapters 5 and 6 introduce text and displayed text. Chapter 5 is especially important because, when you type a LATEX document, most of your time is spent typing text. The topics covered include special characters and accents, hyphenation, fonts, and spacing. Chapter 6 covers displayed text, including lists and tables, and for the mathematician, proclamations (theoremlike structures) and proofs. Typing math is the heart of any mathematical typesetting system. Chapter 7 discusses inline formulas in detail, including basic constructs, delimiters, operators, math accents, and horizontally stretchable lines. The chapter concludes with the Formula Gallery. Math symbols are covered in three sections in Chapter 8. How to space them, how to build new ones. We also look at the closely related subjects of math alphabets and fonts. Then we discuss tagging and grouping equations. LATEX knows a lot about typesetting an inline formula, but not much about how to display a multiline formula. Chapter 9 presents the numerous tools LATEX offers to help you do that. We start with a Visual Guide to help you get oriented. Part III discusses the parts of a LATEX document. In Chapter 10, you learn about the structure of a LATEX document. The most important topics are sectioning and crossreferencing. In Chapter 11, we discuss the amsart document class for articles. In particular, I present the title page information. Chapter 11 also features sampart.tex, a sample article for amsart, first in typeset form, then in mixed form, juxtaposing the source file and the typeset article. You can learn a lot about LATEX just by reading the source file one paragraph at a time and seeing how that paragraph is typeset. We con
xxxii
Introduction clude this chapter with a brief description of the AMS distribution, the packages and document classes, of which amsart is a part. In Chapter 12 the most commonly used legacy document classes are presented, article, report, and letter (the book class is discussed in Chapter 18), along with a description of the standard LATEX distribution. Although article is not as sophisticated as amsart, it is commonly used for articles not meant for publication. In Part IV, we start with Chapter 13, discussing PDF files, hyperlinks, and the hyperref package. This prepares you for presentations, which are PDF files with hyperlinks. In Chapter 14 we utilize the beamer package for making LATEX presentations. Part V (Chapter 15) introduces techniques to customize LATEX: userdefined commands, userdefined environments, and command files. We present a sample command file, newlattice.sty, and a version of the sample article utilizing this command file. You learn how parameters that affect LATEX’s behavior are stored in counters and length commands, how to change them, and how to design your own custom lists. A final section discusses the pitfalls of customization. In Part VI (Chapters 16 and 17), we discuss the special needs of longer documents. Two applications, contained in the standard LATEX distribution, B IBTEX and MakeIndex, make compiling large bibliographies and indexes much easier. LATEX provides the book and the amsbook document classes to serve as foundations for welldesigned books. We discuss these in Chapter 18. Better quality books have to use document classes designed by professionals. We provide some sample pages from a book using Springer’s svmono.cls document class. Detailed instructions are given in Appendix A on how to install LATEX on a PC and a Mac. On a PC we install WinEdt and MiKTeX. On a Mac, we install MacTeX, which consists of TEX Live and TeXShop. For both installations, we describe the editing cycle and three productivity tools in sufficient detail so that you be able to handle the tasks on the sample files of the Short Course. You will probably find yourself referring to Appendices B and C time and again. They contain the math and text symbol tables. Appendix D relates some historical background material on LATEX. It gives you some insight into how LATEX developed and how it works. Appendix E discusses the many ways we can find LATEX material on the Internet. Appendix F is a brief introduction to the use of PostScript fonts in a LATEX document. Appendix G briefly describes the use of LATEX for languages other than American English. Finally, Appendix H discusses what we left out and points you towards some areas for further reading.
xxxiii
Introduction
Mission statement This book is a guide for typesetting mathematical documents within the constraints imposed by LATEX, an elaborate system with hundreds of rules. LATEX allows you to perform almost any mathematical typesetting task through the appropriate application of its rules. You can customize LATEX by introducing userdefined commands and environments and by changing LATEX parameters. You can also extend LATEX by invoking packages that accomplish special tasks. It is not my goal to survey the hundreds of LATEX packages you can utilize to enhance LATEX to teach how to write TEX code and to create your own packages to discuss how to design beautiful documents by writing document classes The definitive book on the first topic is Frank Mittelbach and Michel Goosens’s The LATEX Companion, 2nd edition [46] (with Johannes Braams, David Carlisle, and Chris Rowley). The second and third topics still await authoritative treatment.
Conventions To make this book easy to read, I use some simple conventions: Explanatory text is set in this typeface: Times. Computer Modern typewriter is used to show what you should type, as well as messages from LaTeX. All the characters in this typeface have the same width, making it easy to recognize. I also use Computer Modern typewriter to indicate – Commands (\parbox) – Environments (\align) – Documents (intrart.tex) – Document classes (amsart) – Document class options (draft) – Folders or directories (work) – The names of packages, which are extensions of LATEX (verbatim) When I show you how something looks when typeset, I use Computer Modern, TEX’s standard typeface:
xxxiv
Introduction
I think you find this typeface sufficiently different from the other typefaces I have used. The strokes are much lighter so that you should not have much difficulty recognizing typeset LATEX material. When the typeset material is a separate paragraph or paragraphs, corner brackets in the margin set it off from the rest of the text—unless it is a displayed formula. For explanations in the text, such as Compare iff with iff, typed as iff and if{f}, respectively. the same typefaces are used. Because they are not set off spatially, it may be a little more difficult to see that iff is set in Computer Modern roman (in Times, it looks like this: iff), whereas iff is set in the Computer Modern typewriter typeface. I usually introduce commands with examples, such as \\[22pt] However, it is sometimes necessary to define the syntax of a command more formally. For instance, \\[length ] where length, typeset in Computer Modern typewriter italic font, represents the value you have to supply. Good luck and have fun.
Email:
[email protected] Home page: http://www.maths.umanitoba.ca/homepages/gratzer.html
CHAPTER
1 Your LATEX
Are you sitting in front of your computer, your LATEX implementation up and running? In this chapter we get you ready to tackle this Short Course. When you are done with Part I, you will be ready to start writing your articles in LATEX. If you do not have a LATEX implementation up and running, go to Appendix A. There you find precise and detailed instructions how to set up LATEX on a PC or a Mac. There is enough in the appendix for you to be able to handle the tasks in this Short Course. You will be pleasantly surprised at how little time it takes to set LATEX up. If you use some variant of UNIX, turn to a UNIX guru who can help you set up LATEX on your computer and guide you through the basics. If all else fails, read the documentation for your UNIX system.
1.1
Your computer We assume very little, only that you are familiar with your keyboard and with the operating system on your computer. You should know standard PC and Mac menus, pull down menus, buttons, tabs, the menu items, such as Edit>Paste, the menu item Paste on the menu Edit. You should understand folders (we use this terminology regardless of the platform, with apologies to our UNIX readers), and you need to know how to save a file and copy a file from one folder to another.
4
Chapter 1
Your LATEX
On a PC, work\test refers to the subfolder test of the folder work. On a Mac, work/test designates this subfolder. To avoid having to write every subfolder twice, we use work/test, with apologies to our PC readers.
1.2
Sample files We work with a few sample documents in this Short Course. You can type the sample documents as presented in the text, or you can download them from the Internet (see Section E.1). The samples folder also contains a copy of SymbolTables.pdf, a PDF version of Appendices B and C, the symbol tables. I suggest you create a folder on your computer named samples, to store the downloaded sample files, and another folder called work, where you will keep your working files. Copy the documents from the samples to the work folder as needed. In this book, the samples and work folders refer to the folders you have created. If you Save As... a sample file under a different name, remember the naming rule.
Rule
Naming of source files The name of a LATEX source file should be one word (no spaces, no special characters), and end with .tex. So first art.tex is bad, but art1.tex and FirstArt.tex are good.
1.3
Editing cycle Watch a friend type a mathematical article in LATEX and you learn some basic steps. 1. A text editor is used to create a LATEX source file. A source file might look like the top window in Figure 1.1: \documentclass{amsart} \begin{document} The hypotenuse: $\sqrt{a^{2} + b^{2}}$. I can type math! \end{document} Note that the source file is different from a typical word processor file. All characters are displayed in the same font and size. 2. Your friend “typesets” the source file (tells the application to produce a typeset version) and views the result on the monitor (the two corners indicate material typeset by LATEX): The hypotenuse:
√
a2 + b2 . I can type math!
1.4
Three productivity tools
5
as in the middle window in Figure 1.1. 3. The editing cycle continues. Your friend goes back and forth between the source file and the typeset version, making changes and observing the results of these changes. 4. The file is printed. Once the typeset version is satisfactory, it is printed, creating a paper version of the typeset article. Alternatively, your friend creates a PDF file of the typeset version (see Chapter 13.1.2). If LATEX finds a mistake when typesetting the source file, it opens a new window, the log window, illustrated as the bottom window in Figure 1.1, and displays an error message. The same message is saved into a file, called the log file. Look at the figures in Appendix A, depicting a variety of editing windows, windows for the typeset article, and log windows for the two LATEX implementations discussed there. Various LATEX implementations have different names for the source file, the text editor, the typeset file, the typeset window, the log window, and the log file. Become familiar with these names for the LATEX implementation you use, so you can follow along with our discussions. In Appendix A, we bring you up to speed for the LATEX implementations discussed therein.
1.4
Three productivity tools Most LATEX implementations have these important productivity tools: Synchronization To move quickly between the source file and the typeset file, most LATEX implementations offer synchronization, the ability to jump from the typeset
Figure 1.1: Windows for the source and typeset files and the log window.
6
Chapter 1
Your LATEX
file to the corresponding place in the source file and from the source file to the corresponding place in the typeset file. Block comment Block comments are very useful: 1. When looking for a LATEX error, you may want LATEX to ignore a block of text in the source file (see page 51). 2. Often you may want to make comments about your project but not have them printed or you may want to keep text on hand while you try a different option. To accomplish this, insert a comment character, %, at the start of each line where the text appears. These lines are ignored when the LATEX file is processed. Select a number of lines in a source document, then by choosing a menu option all the lines (the whole block) are commented out (a % sign is placed at the beginning of each line). This is block comment. The reverse is block uncomment. Jump to a line This is specified by the line number in the source file. To find an error, LATEX suggests that you jump to a line. Find out how your LATEX implements these features. In Appendix A, we discuss how these features are implemented for the LATEX we install. Pay careful attention how your LATEX implementation works. This enables you to rapidly perform the editing cycle and utilize the productivity tools when necessary.
CHAPTER
2 Typing text
In this chapter, I introduce you to typesetting text by working through examples. More details are provided throughout the book, in particular, in Chapters 5 and 6. A source file is made up of text, math (formulas), and instructions (commands) to LATEX. For instance, consider the following variant of the first sentence of this paragraph: A source file is made up of text, math (e.g., $\sqrt{5}$), and \emph{instructions to} \LaTeX. This typesets as A source file is made up of text, math (e.g.,
√
5), and instructions to LATEX.
In this sentence, the first part A source file is made up of text, math (e.g., is text. Then $\sqrt{5}$
8
Chapter 2
Typing text
is math ), and is text again. Finally, \emph{instructions to} \LaTeX. are instructions. The instruction \emph is a command with an argument, while the instruction \LaTeX is a command without an argument. Commands, as a rule, start with a backslash ( \ ) and tell LATEX to do something special. In this case, the command \emph emphasizes its argument (the text between the braces). Another kind of instruction to LATEX is called an environment. For instance, the commands \begin{flushright} and \end{flushright} enclose a flushright environment; the content, that is, the text that is typed between these two commands, is right justified (lined up against the right margin) when typeset. (The flushleft environment creates left justified text; the center environment creates text that is centered horizontally on the page.) In practice, text, math, and instructions (commands) are mixed. For example, My first integral: $\int \zeta^{2}(x) \, dx$. is a mixture of all three; it typesets as My first integral:
R
ζ 2 (x) dx.
Creating a document in LATEX requires that we type the text and math in the source file. So we start with the keyboard, proceed to type a short note, and learn some simple rules for typing text in LATEX.
2.1
The keyboard The following keys are used to type text in a source file: az
AZ
09
+ = * / ( ) [ ]
2.2
9
Your first note You may also use the following punctuation marks: ,
;
.
?
!
:
‘
’

and the space bar, the Tab key, and the Return (or Enter) key. Since TEX source files are “pure text” (ASCII files), they are very portable. There is one possible problem limiting this portability, the line endings used in the source file. When you press the Return key, your text editor writes an invisible code into your source file that indicates where the line ends. Since this code may be different on different platforms (PC, Mac, and UNIX), you may have problems reading a source file created on a different platform. Luckily, many text editors include the ability to switch endofline codes and some, including the editors in WinEdt and TeXShop, do so automatically. Finally, there are thirteen special keys that are mostly used in LATEX commands: #
$
%
&
~
_
^
\
{
}
@
"

If you need to have these characters typeset in your document, there are commands to produce them. For instance, $ is typed as \$, the underscore, , is typed as \_, and % is typed as \%. Only @ requires no special command, type @ to print @. There are also commands to produce composite characters, such as accented characters, for example ¨ a, which is typed as \"{a}. See Section 5.4.4 for a complete discussion of symbols not available directly from the keyboard and Appendix C for the text symbol tables. Appendices B and C are reproduced in the samples folder as a PDF file, SymbolTables.pdf. LATEX prohibits the use of other keys on your keyboard—unless you are using a version of LATEX that is set up to work with nonEnglish languages (see Appendix G). When trying to typeset a source file that contains a prohibited character, LATEX displays an error message similar to the following: ! Text line contains an invalid character. l.222 completely irreducible^^? ^^? In this message, l.222 means line 222 of your source file. You must edit that line to remove the character that LATEX cannot understand. The log file (see Section D.3.4) also contains this message. For more about LATEX error messages, see Sections 3.2 and 4.3.1.
2.2
Your first note We start our discussion on how to type a note in LATEX with a simple example. Suppose you want to use LATEX to produce the following:
10
Chapter 2
Typing text
It is of some concern to me that the terminology used in multisection math courses is not uniform. In several sections of the course on matrix theory, the term “hamiltonianreduced” is used. I, personally, would rather call these “hypersimple”. I invite others to comment on this problem. Of special concern to me is the terminology in the course by Prof. Rudi Hochschwabauer. Since his field is new, there is no accepted terminology. It is imperative that we arrive at a satisfactory solution. To produce this typeset document, create a new file in your work folder with the name note1.tex. Type the following, including the spacing and linebreaks shown, but not the line numbers: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
% Sample file: note1.tex \documentclass{sample} \begin{document} It is of some concern to me that the terminology used in multisection math courses is not uniform. In several sections of the course on matrix theory, the term ‘‘hamiltonianreduced’’ is used. I, personally, would rather call these ‘‘hypersimple’’. I invite others to comment on this problem. Of special concern to me is the terminology in the course by Prof.~Rudi Hochschwabauer. Since his field is new, there is no accepted terminology. It is imperative that we arrive at a satisfactory solution. \end{document}
Alternatively, copy the note1.tex file from the samples folder (see page 4). Make sure that sample.cls is in your work folder. The first line of note1.tex starts with %. Such lines are called comments and are ignored by LATEX. Commenting is very useful. For example, if you want to add some notes to your source file and you do not want those notes to appear in the typeset version of your article, you can begin those lines with a %. You can also comment out part of a line:
2.2
Your first note
11
simply put, we believe % actually, it’s not so simple Everything on the line after the % character is ignored by LATEX. Line 2 specifies the document class (in our case, sample)1 that controls how the document is formatted. The text of the note is typed within the document environment, that is, between the lines \begin{document} and \end{document} Now typeset note1.tex. If you use WinEdt, click on the TeXify icon. If you use TeXShop, click the Typeset button. You should get the typeset document as shown on page 10. As you can see from this example, LATEX is different from a word processor. It disregards the way you input and position the text, and follows only the formatting instructions given by the markup commands. LATEX notices when you put a blank space in the text, but it ignores how many blank spaces have been inserted. LATEX does not distinguish between a blank space (hitting the space bar), a tab (hitting the Tab key), and a single carriage return (hitting Return once). However, hitting Return twice gives a blank line; one or more blank lines mark the end of a paragraph. LATEX, by default, fully justifies text by placing a flexible amount of space between words—the interword space—and a somewhat larger space between sentences—the intersentence space. If you have to force an interword space, you can use the \ command (in LATEX books, we use the symbol to mean a blank space). See Section 5.2.2 for a full discussion. The ~ (tilde) command also forces an interword space, but with a difference; it keeps the words on the same line. This command is called a tie or nonbreakable space (see Section 5.4.3). Note that on lines 11 and 13, the left double quotes are typed as ‘‘ (two left single quotes) and the right double quotes are typed as ’’ (two right single quotes or apostrophes). The left single quote key is not always easy to find. On an American keyboard,2 it is usually hidden in the upperleft or upperright corner of the keyboard, and shares a key with the tilde (~). 1 I know you have never heard of the sample document class. It is a special class created for these exercises. You can find it in the samples folder (see page 4). If you have not yet copied it over to the work folder, do so now. 2 The location of special keys on the keyboard depends on the country where the computer was sold. It also depends on whether the computer is a PC or a Mac. In addition, notebooks tend to have fewer keys than desktop computers. Fun assignment: Find the tilde (~) on a Spanish and on a Hungarian keyboard.
12
2.3
Chapter 2
Typing text
Lines too wide LATEX reads the text in the source file one line at a time and when the end of a paragraph is reached, LATEX typesets the entire paragraph. Occasionally, LATEX gets into trouble when trying to split the paragraph into typeset lines. To illustrate this situation, modify note1.tex. In the second sentence, replace term by strange term and in the fourth sentence, delete Rudi , including the blank space following Rudi. Now save this modified file in your work folder using the name note1b.tex. You can also find note1b.tex in the samples folder (see page 4). Typesetting note1b.tex, you obtain the following: It is of some concern to me that the terminology used in multisection math courses is not uniform. In several sections of the course on matrix theory, the strange term “hamiltonianreduced” is used. I, personally, would rather call these “hypersimple”. I invite others to comment on this problem. Of special concern to me is the terminology in the course by Prof. Hochschwabauer. Since his field is new, there is no accepted terminology. It is imperative that we arrive at a satisfactory solution. The first line of paragraph two is about 1/4 inch too wide. The first line of paragraph three is even wider. In the log window, LATEX displays the following messages: Overfull \hbox (15.38948pt too wide) in paragraph at lines 915 []\OT1/cmr/m/n/10 In several sections of the course on matrix theory, the strange term ‘‘hamiltonianOverfull \hbox (23.27834pt too wide) in paragraph at lines 1621 []\OT1/cmr/m/n/10 Of special concern to me is the terminology in the course by Prof. Hochschwabauer. You will find the same messages in the log file (see Sections 1.3 and D.2.1). The first message, Overfull \hbox (15.38948pt too wide) in paragraph at lines 915 refers to the second paragraph (lines 9–15 in the source file—its location in the typeset document is not specified). The typeset version of this paragraph has a line that is 15.38948 points too wide. LATEX uses points (pt) to measure distances; there are about 72 points in 1 inch (or about 28 points in 1 cm).
2.4
13
More text features The next two lines, []\OT1/cmr/m/n/10 In several sections of the course on matrix theory, the strange term ‘‘hamiltonianidentify the source of the problem: LATEX did not properly hyphenate the word hamiltonianreduced because it (automatically) hyphenates a hyphenated word only at the hyphen. The second reference, Overfull \hbox (23.27834pt too wide) in paragraph at lines 1621
is to the third paragraph (lines 16–21 of the source file). There is a problem with the word Hochschwabauer; LATEX’s standard hyphenation routine cannot handle it (a German hyphenation routine would have no difficulty hyphenating this name—see Appendix G). If you encounter such a problem, you can either try to reword the sentence or insert one or more optional (or discretionary) hyphen commands (\), which tell LATEX where it may hyphenate the word. In this case, you can rewrite Hochschwabauer as Hoch\schwa\bauer and the second hyphenation problem disappears. You can also utilize the \hyphenation command (see Section 5.4.9). Sometimes a small horizontal overflow can be difficult to spot. The draft document class option may help (see Sections 11.5, 12.1.2, and 18.1 for more about document class options). LATEX places a black box (or slug) in the margin to mark an overfull line. You can invoke this option by changing the \documentclass line to \documentclass[draft]{sample} A version of note1b.tex with this option can be found in the samples folder under the name noteslug.tex. Typeset it to see the “slugs”.
2.4
More text features Next, we produce the following note: September 12, 2006 From the desk of George Gr¨atzer October 7–21 please use my temporary email address: George
[email protected] 14
Chapter 2
Typing text
Type in the source file, without the line numbers. Save it as note2.tex in your work folder (note2.tex can be found in the samples folder—see page 4): 1 2 3 4 5 6 7 8 9 10 11 12 13 14
% Sample file: note2.tex \documentclass{sample} \begin{document} \begin{flushright} \today \end{flushright} \textbf{From the desk of George Gr\"{a}tzer}\\[22pt] October~721 \emph{please} use my temporary email address: \begin{center} \texttt{George\
[email protected]} \end{center} \end{document} This note introduces several additional text features of LATEX:
The \today command (in line 6) to display the date on which the document is typeset (so you will see a date different from the date shown above in your own typeset document). The environments to right justify (lines 5–7) and center (lines 1113) text. The commands to change the text style, including the \emph command (line 8) to emphasize text, the \textbf command (line 9) for bold text, and the \texttt command (line 12) to produce typewriter style text. These are commands with arguments. In each case, the argument of the command follows the name of the command and is typed between braces, that is, between { and }. The form of the LATEX commands: Almost all LATEX commands start with a backslash ( \ ) followed by the command name. For instance, \textbf is a command and textbf is the command name. The command name is terminated by the first nonalphabetic character, that is, by any character other than a–z or A–Z. So textbf1 is not a command name, in fact, \textbf1 typesets as 1. (Let us look at this a bit more closely. \textbf is a valid command. If a command needs an argument and is not followed by braces, then it takes the next character as its argument. So \textbf1 is the command \textbf with the argument 1, which typesets as bold 1: 1.) Note that command names are case sensitive. Typing \Textbf or \TEXTBF generates an error message. The multiple role of hyphens: Double hyphens are used for number ranges. For example, 721 (in line 9) typesets as 7–21. The punctuation mark – is called an en
2.4
More text features
15
dash. Use triple hyphens for the em dash punctuation mark—such as the one in this sentence. The new line command, \\ (or \newline): To create additional space between lines (as in the last note, under the line From the desk. . . ), you can use the \\ command and specify an appropriate amount of vertical space: \\[22pt]. Note that this command uses square brackets rather than braces because the argument is optional. The distance may be given in points (pt), centimeters (cm), or inches (in). (There is an analogous new page command, \newpage, not used in this short note.) Special rules for special characters (see Section 2.1), for accented characters and for some European characters. For instance, the accented character ¨a is typed as \"{a}. Accents are explained in Section 5.4.7 (see also the tables in Section C.2). When you need to know more about typing text than we have discussed here, see Chapters 5 and 6. See also Appendix C, where all text symbols are organized into tables. Recall that we also have the SymbolTables.pdf in the samples folder.
CHAPTER
3 Typing math
While marking up text in LATEX is easy, marking up math is less intuitive because math formulas are twodimensional constructs and we have to mark them up with a onedimensional string of characters. However, even the most complicated twodimensional formula is made up of fairly simple building blocks. So by concentrating on the building blocks—selectively, just learn the ones you need—you can get started with math quickly.
3.1
A note with math In addition to the regular text keys and the 13 special keys discussed in Section 2.1, two more keys are used to type math:
The formula 2 < x > y (typed as $2 < x >y$) uses both. Note that such math formulas, called inline, are enclosed by $ symbols. We discuss shortly another kind of math formula called displayed. We begin typesetting math with the following note:
18
Chapter 3
Typing math
In firstyear calculus, we define intervals such as (u, v) and (u, ∞). Such an interval is a neighborhood of a if a is in the interval. Students should realize that ∞ is only a symbol, not a number. This is important since we soon introduce concepts such as limx→∞ f (x). When we introduce the derivative lim
x→a
f (x) − f (a) , x−a
we assume that the function is defined and continuous in a neighborhood of a. To create the source file for this mixed text and math note, create a new document with your text editor. Name it math.tex, place it in the work folder, and type in the following source file—without the line numbers—or simply copy math.tex from the samples folder (see page 4): 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
% Sample file: math.tex \documentclass{sample} \begin{document} In firstyear calculus, we define intervals such as $(u, v)$ and $(u, \infty)$. Such an interval is a \emph{neighborhood} of $a$ if $a$ is in the interval. Students should realize that $\infty$ is only a symbol, not a number. This is important since we soon introduce concepts such as $\lim_{x \to \infty} f(x)$. When we introduce the derivative \[ \lim_{x \to a} \frac{f(x)  f(a)}{x  a}, \] we assume that the function is defined and continuous in a neighborhood of $a$. \end{document}
This note introduces several basic concepts of math in LATEX: There are two kinds of math formulas and environments in math.tex: – Inline math environments open and close with $ (as seen throughout this book) or open with \( and close with \). – Displayed math environments open with \[ and close with \].
3.2
19
Errors in math
Within math environments, LATEX uses its own spacing rules and completely ignores the white space you type, with two exceptions: – Spaces that terminate commands. So in $\infty a$ the space is not ignored, $\inftya$ produces an error. – Spaces in the arguments of commands that temporarily revert to regular text. \text is such a command (see Sections 3.3 and 7.4.6). The white space that you add when typing math is important only for the readability of the source file. We summarize with a simple rule.
Rule
Spacing in text and math Many spaces equal one space in text, whereas your spacing is ignored in math, unless the space terminates a command. A math symbol is invoked by a command. For example, the command for ∞ is \infty and the command for → is \to. The math symbols are organized into tables in Appendix B (see also SymbolTables.pdf in the samples folder). Some commands, such as \sqrt, need arguments enclosed by { and }. To typeset √ 5, type $\sqrt{5}$, where \sqrt is the command and 5 is the argument. Some commands need more than one argument. To get 3+x 5 type \[
\frac{3+x}{5}
\] where \frac is the command, 3+x and 5 are the arguments—we indent for readability.
3.2
Errors in math Even in such a simple note there are opportunities for errors. To help familiarize yourself with some of the most commonly seen LATEX math errors and their causes, we deliberately introduce mistakes into math.tex. The version of math.tex with mistakes is mathb.tex. By inserting and deleting % signs, you make the mistakes visible to LATEX one at a time—recall that lines starting with % are comments and are therefore ignored by LATEX.
20
Chapter 3
Typing math
Type the following source file, and save it under the name mathb.tex in the work folder or copy the file mathb.tex from the samples folder (see page 4). Do not type the line numbers—they are shown here to help you with the experiments. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
% Sample file: mathb.tex \documentclass{sample} \begin{document} In firstyear calculus, we define intervals such % as $(u, v)$ and $(u, \infty)$. Such an interval as $(u, v)$ and (u, \infty)$. Such an interval is a \emph{neighborhood} of $a$ if $a$ is in the interval. Students should realize that $\infty$ is only a symbol, not a number. This is important since we soon introduce concepts such as $\lim_{x \to \infty} f(x)$. %such as $\lim_{x \to \infty f(x)$. When we introduce the derivative \[ \lim_{x \to a} \frac{f(x)  f(a)}{x  a} % \lim_{x \to a} \frac{f(x)  f(a) x  a} \] we assume that the function is defined and continuous in a neighborhood of $a$. \end{document}
Experiment 1 In line 7, the $ before the (u is missing. Typeset the mathb.tex source file. LATEX generates the following error message: ! Missing $ inserted. $ l.7 as $(u, v)$ and (u, \infty )$. Such an interval Since the $ was omitted, LATEX reads (u, \infty) as text; but the \infty command instructs LATEX to typeset a math symbol, which can only be done in a math formula. So LATEX offers to put a $ in front of \infty while typesetting the source file—it does not put the $ in the source file itself. LATEX attempts a cure, but in this example it comes too late, because the math formula should start just before (u.
3.2
21
Errors in math
Whenever you see the ? prompt, you may press Return to ignore the error and continue typesetting the document. Section D.4 lists a number of other options and prompts. Experiment 2 Uncomment line 6 by deleting the % at the beginning of line 6 and comment out line 7 by inserting a % at the beginning of line 7. This eliminates the previous error. Uncomment line 14 and comment out line 13. This introduces a new error, the closing brace of the subscript is missing. Now typeset the note. You get the error message ! Missing } inserted. } l.14 such as $\lim_{x \to \infty f(x)$ LATEX reports that a closing brace (}) is missing, but it is not sure where the brace should be. LATEX noticed that a subscript (see page 23) started with {, but LATEX reached the end of the math formula before finding a closing brace }. To remedy this, you must look in the formula for an opening brace { that is not balanced, and insert the missing closing brace }. Make the necessary change and typeset again to view the difference. Experiment 3 Uncomment line 13 and comment out line 14, removing the previous error. Delete the % at the beginning of line 19 and insert a % at the beginning of line 18, introducing our final error, omitting the closing brace of the first argument and the opening brace of the second argument of \frac. Save and typeset the file. You get the error message ! Too many }’s. \frac #1#2>{\begingroup #1\endgroup \@@over #2} l.20 \] LATEX got confused. The second line of the message explains that \frac has two arguments and it is not working out, but the error is incorrectly identified. If the typo on line 19 is \frac{f(x)f(a) xa}g, then LATEX produces f (x) − f (a)x − a g and no error message is generated. Experiment 4 Make sure all the errors are commented out. Typeset mathb.tex, testing that there are no errors. Now delete the two $ signs in line 22, that is, replace $a$ by a. Typeset the file. It typesets with no errors. Here is the last line of the typeset file you get:
22
Chapter 3
Typing math
we assume that the function is defined and continuous in a neighborhood of a. instead of we assume that the function is defined and continuous in a neighborhood of a. This is probably the error most often made by beginners. There is no error message by LATEX and the typeset version looks good. You need sharp eyes to catch such an error. See Section 4.3.1 for more information about finding and fixing problems in your LATEX source files.
3.3
Building blocks of a formula A formula is built from a large collection of components. We group them as follows: Arithmetic – Subscripts and superscripts Binomial coefficients Congruences Delimiters Ellipses Integrals Math accents Matrices Operators – Large operators Roots Text In this section, I describe each of these groups, and provide examples illustrating their use. Arithmetic The arithmetic operations a + b, a − b, −a, a/b, and ab are typed in the natural way (the spaces are typed only for readability, others may type fewer spaces):
3.3
23
Building blocks of a formula $a + b$, $a  b$, $a$, $a / b$, $a b$
If you wish to use · or × for multiplication, as in a · b or a × b, use \cdot or \times, respectively. The expressions a · b and a × b are typed as follows: $a \cdot b$
$a \times b$
Displayed fractions, such as
1 + 2x x + y + xy
are typed with \frac: \[ \]
\frac{1 + 2x}{x + y + xy}
The \frac command is seldom used inline because it can disrupt the interline spacing of the paragraph. Subscripts and superscripts Subscripts are typed with (underscore) and superscripts with ^ (caret). Subscripts and superscripts should be enclosed in braces, that is, typed between { and }. To get a1 , type $a_{1}$. Omitting the braces in this example causes no harm, but to get a10 , you must type $a_{10}$. Indeed, $a_10$ is typeset as a1 0. Further examples, ai1 , a2 , ai1 , a2n , are typed as $a_{i_{1}}$, $a^{2}$, $a^{i_{1}}$, $a_{n}^{2}$ There is one symbol, the prime (’), that is automatically superscripted in math. To get f 0 (x), just type $f’(x)$. Binomial coefficients Binomial coefficients are typeset with the \binom command. a For example, b+c is typed inline as $\binom{a}{b + c}$ whereas a displayed version,
n2 −1 2
n+1 is typed as \[ \]
\binom{ \frac{n^{2}  1}{2} }{n + 1}
24
Chapter 3
Typing math
Congruences The two most important forms are a ≡ v (mod θ) a ≡ v (θ)
typed as typed as
$a \equiv v \pmod{\theta}$ $a \equiv v \pod{\theta}$
Delimiters Parentheses and square brackets are examples of delimiters. They are used to delimit some subformulas, as in $[(a*b)+(c*d)]^{2}$, which typesets as [(a ∗ b) + (c ∗ d)]2 . LATEX can be instructed to expand them vertically to enclose a formula such as 2 1+x 2 + y2 which is typed as \[ \]
\left( \frac{1 + x}{2 + y^{2}} \right)^{2}
The \left( and \right) commands tell LATEX to size the parentheses correctly, relative to the size of the symbols inside the parentheses. Two further examples, a + b
2
A 2 , would be typed as \[ \left \frac{a + b}{2} \right, \quad \left\ A^{2} \right\ \] where \quad is a spacing command (see Sections 8.1 and B.9). Additional delimiters are listed in Sections 7.5 and B.6. Ellipses The ellipsis ( . . . ) in text is provided by the \dots command: A. . . Z is typed as A \dots Z In formulas, the ellipsis is printed either as low (or ontheline) dots: F (x1 , . . . , xn ) is typed as $F(x_{1}, \dots, x_{n})$ or as centered dots: x1 + · · · + xn
is typed as $x_{1} + \dots + x_{n}$
The command \dots typesets the correct ellipsis with the correct spacing in most cases. If it does not, see Section 7.4.3 on how to specify the appropriate ellipsis from the four types available.
3.3
25
Building blocks of a formula
Integrals The command for an integral is \int. The lower limit is specified as a subscriptR and the upper limit is specified as a superscript. For example, the π formula 0 sin x dx = 2 is typed as $\int_{0}^{\pi} \sin x \, dx = 2$ where \, is a spacing command (see Sections 8.1 and B.9). Math accents The four most frequently used math accents are: a ¯
typed as
$\bar{a}$
a ˆ
typed as
$\hat{a}$
a ˜
typed as
$\tilde{a}$
~a
typed as
$\vec{a}$
See Sections 7.7 and B.8 for complete lists. Matrices You type the matrix a+b+c uv a+b u+v
x−y z
27 134
with the \matrix command \[ \begin{matrix} a + b + c & uv & x  y & 27\\ a + b & u + v & z & 134 \end{matrix} \] The matrix environment separates adjacent matrix elements within a row with ampersands (&). Rows are separated by new line commands (\\). Do not end the last row with a new line command! The matrix environment has to appear within a math environment, as in the example. As a rule, it is in a displayed math environment, since inline it appears too large. It can be used in the align environment discussed in Section 3.4.2. The matrix environment does not provide delimiters. Several variants do, including pmatrix and vmatrix. For example, 30 7 a+b+c uv A= a+b u + v 3 17 is typed as follows:
26
Chapter 3 \[
\]
Typing math
\mathbf{A} = \begin{pmatrix} a + b + c & uv\\ a + b & u + v \end{pmatrix} \begin{vmatrix} 30 & 7\\ 3 & 17 \end{vmatrix}
As you can see, pmatrix typesets as a matrix between a pair of \left( and \right) commands, while vmatrix typesets as a matrix between a pair of \left and \right commands. See Section 9.7.1 for a listing of all the matrix variants. Operators To typeset the sine function, sin x, type $\sin x$. Note that $sin x$ would be typeset as sinx because LATEX interprets this expression as the product of four variables. LATEX calls \sin an operator. Sections 7.6.1 and B.7 list a number of operators. See Section 7.6.2 for userdefined operators. Some are just like \sin. Others produce a more complex display, for example, lim f (x) = 0
x→0
is typed as \[ \]
\lim_{x \to 0} f(x) = 0
Large operators The command for sum is \sum and for product is \prod. The following examples, n n X Y x2i x2i i=1
i=1
are typed as \[
\]
\sum_{i=1}^{n} x_{i}^{2} \qquad \prod_{i=1}^{n} x_{i}^{2}
3.4
27
Displayed formulas
where \qquad is a spacing command (see Sections 8.1 and B.9) used to separate the two formulas, yielding twice the space produced by \quad. Sums and products are examples of large operators. They are typeset bigger when displayed than inline. They are listed in Sections 7.6.4 and B.7.1. √ Roots \sqrt produces a square root. For instance, a + 2b is typed as $\sqrt{a + 2b}$ √ The nth root, n 5, requires the use of an optional argument, which is specified using brackets (see Section 5.3.1): $\sqrt[n]{5}$. Text You can include text in a formula with a \text command. For instance, a = b,
by assumption
is typed as \[ \]
a = b, \text{\qquad by assumption}
Note the spacing command \qquad in the argument of \text. You could also type \[
a = b, \qquad \text{by assumption}
\] because \qquad works in math as well as in text (see Sections 8.1 and B.9).
3.4
Displayed formulas
3.4.1 Equations The equation environment creates a displayed math formula and automatically generates an equation number. The equation Z π (1) sin x dx = 2 0
is typed as \begin{equation}\label{E:firstInt} \int_{0}^{\pi} \sin x \, dx = 2 \end{equation}
28
Chapter 3
Typing math
The equation number, which is automatically generated, depends on how many other numbered equations occur before the given equation. To reference this formula without having to remember a number—which may change when you edit your document—give the equation a symbolic label by using the \label command and refer to the equation in your document by using the symbolic label, the argument of the \label command. In this example, I have called the first equation firstInt (first integral), and used the convention that the label of an equation starts with E:, so that the complete \label command is \label{E:firstInt} The number of this formula is referenced with the \ref command. Its page is referenced using the \pageref command. For example, to get see (1) on page 27 type see~(\ref{E:firstInt}) on page~\pageref{E:firstInt} The \eqref command provides the reference number in parentheses. So the last example could be typed see~\eqref{E:firstInt} on page~\pageref{E:firstInt} The \eqref command is smart. Even if the equation number is referenced in emphasized or italicized text, the reference typesets upright (in roman type). Note the use of the nonbreakable space (~) to ensure that when typeset the equation number is on the same line as the word see. (See the footnote on page 11.) You should always use a nonbreakable space to link a \ref command to the name of its part, for instance, equation, page, section, chapter. Use two nonbreakable spaces in Sections~\ref{S:main} and~\ref{S:subsidiary}. The main advantage of this crossreferencing system is that when you add, delete, or rearrange equations, LATEX automatically renumbers the equations and adjusts the references that appear in your typeset document. You can split a long article into two or move a section to the end, and LATEX takes care of the renumbering. This significantly reduces the amount of time you need to spend working on your document. It also reduces the potential for errors in the finished project.
Rule
Typeset twice For renumbering to work, you have to typeset the source file twice.
3.4
29
Displayed formulas
The first run creates a list of references that need to be linked. The second creates the cross references and inserts the relevant text throughout the document (see Sections 18.2 and D.3.4). LATEX issues a warning if you forget. Such warnings do not interrupt the typesetting, you only see them in the log window—if the window is visible—and in the log file. It is a good idea to check for warnings periodically. An equation is numbered whether or not there is a \label command attached to it. Of course, if there is no \label command, the number generated for the equation by LATEX cannot be referenced with the command \ref or \eqref. The system described here is called symbolic referencing. The symbol for the number is the argument of the \label command, and that symbol can be referenced with \ref, \eqref, or \pageref commands. LATEX uses the same mechanism for all of the generated numbering systems: sections, subsections, subsubsections, equations, theorems, lemmas, and bibliographic references—except that for bibliographic references, LATEX uses the \bibitem command to define a bibliographic item and the \cite command to cite a bibliographic item (see Section 4.2.4 and Chapter 16). What happens if you misspell a reference, e.g., typing \ref{E:firstint} instead of \ref{E:firstInt}? LATEX typesets ??. There are two warnings in the log file: LaTeX Warning: Reference ‘E:firstint’ on page 39 undefined on input line 475. for the typeset page and the other one close to the end: LaTeX Warning: There were undefined references. If a \cite is misspelled, you get [?] and similar warnings. Equations can also be tagged by attaching a name to the formula with the \tag command. The tag replaces the equation number. For example, Z (Int)
0
π
sin x dx = 2
is typed as \begin{equation} \int_{0}^{\pi} \sin x \, dx = 2 \tag{Int} \end{equation} Tags (of the type discussed here) are absolute. This equation is always referred to as (Int). Equation numbers, on the other hand, are relative, they may change when equations are added, deleted, or rearranged.
30
3.4.2
Chapter 3
Typing math
Aligned formulas LATEX has many ways to typeset multiline formulas. We discuss three constructs in this section: simple alignment, annotated alignment, and cases. See Chapter 9 for many others. For simple and annotated alignment we use the align environment. Each line in the align environment is a separate equation, which LATEX automatically numbers. Simple alignment Simple alignment is used to align two or more formulas. To obtain the formulas (2) (3) (4)
r2 = s2 + t2 , 2u + 1 = v + wα , y+z x= √ ; s + 2u
type the following, using \\ as the line separator and & as the alignment point: \begin{align} r^{2} &= s^{2} + t^{2}, \label{E:Pyth}\\ 2u + 1 &= v + w^{\alpha}, \label{E:alpha}\\ x &= \frac{y + z}{\sqrt{s + 2u}};\label{E:frac} \end{align} Note that you should not have a \\ to terminate the last line. Figure 3.1 displays the source and the typeset versions of formulas (2)–(4), emphasizing the alignment points of the source and the typeset formula. Of course, in the source, the alignment points do not have to line up. These formulas are numbered (2), (3), and (4) because they are preceded by one numbered equation earlier in this section. The align environment can also be used to break a long formula into two or more parts. Since numbering both lines in such a case would be undesirable, you can prevent the numbering of the second line by using the \notag command in the second part of the formula. For example, Z f (x) + g(x) 1 + f (x)g(x) √ (5) h(x) = + dx 1 + f 2 (x) 1 − sin x Z 1 + f (x) = dx − 2 tan−1 (x − 2) 1 + g(x) is typed as follows: \begin{align} h(x) &= \int \left( \frac{f(x) + g(x)}{1+ f^{2}(x)}
3.4
31
Displayed formulas
r^{2} &= s^{2} + t^{2}, 2u + 1 &= v + w^{\alpha}, x &= \frac{y + z}{\sqrt{s + 2u}};
\label{E:Pyth}\\ \label{E:alpha}\\ \label{E:frac}
alignment points of formulas (2) (3) (4)
r2 = s2 + t2 , 2u + 1 = v + wα , y+z x= √ ; s + 2u
Figure 3.1: Simple alignment: source and typeset.
+ \frac{1+ f(x)g(x)}{\sqrt{1  \sin x}} \right) \, dx\label{E:longInt}\\ &= \int \frac{1 + f(x)}{1 + g(x) } \, dx  2 \tan^{1}(x2)\notag \end{align}
The rules for simple alignment are easy to remember.
Rule
Simple alignments Use the align environment. Separate the lines with \\. In each line, indicate the alignment point with &, one & per line. If the alignment point is adjacent to an =, +, and so on, place it before to ensure proper spacing. Place a \notag command in each line that you do not wish numbered. If no line should be numbered, use the align* environment. Place a \label command in each numbered line you may want to reference with \ref, \eqref, or \pageref.
32
Chapter 3
aligned formulas x &= x \wedge (y \vee z) &= (x \wedge y) \vee (x \wedge z) &= y \vee z.
Typing math
annotation & &\text{(by distributivity)}\\ & &\text{(by condition (M))}\\
alignment points of annotations
alignment points of formulas aligned formulas x = x ∧ (y ∨ z) = (x ∧ y) ∨ (x ∧ z)
annotation (by distributivity) (by condition (M))
= y ∨ z. alignment points of formulas
alignment points of annotations
Figure 3.2: Annotated alignment: source and typeset. Annotated alignment Annotated alignment allows you to align formulas and their annotations, that is, explanatory text, separately (see Figure 3.2): (6)
x = x ∧ (y ∨ z) = (x ∧ y) ∨ (x ∧ z)
(by distributivity) (by condition (M))
= y ∨ z. This example is typed as \begin{align} x &= x \wedge (y \vee z) &&\text{(by distributivity)}\label{E:DoAlign}\\ &= (x \wedge y) \vee (x \wedge z) &&\text{(by condition (M))} \notag\\ &= y \vee z. \notag \end{align} The rules for annotated alignment are similar to the rules of simple alignment. In each line, in addition to the alignment point marked by &, there is also a mark for the start of the annotation: &&.
3.4
33
Displayed formulas
3.4.3 Cases The cases construct is a specialized matrix. It has to appear within a math environment such as the equation environment (see Section 3.4.1) or the align environment (see Section 3.4.2). Here is a typical example: 2 if x < 0; −x , f (x) = α + x, if 0 ≤ x ≤ 1; x 2 , otherwise. is typed as follows: \[
\]
f(x)= \begin{cases} x^{2}, \alpha + x, x^{2}, \end{cases}
&\text{if $x < 0$;}\\ &\text{if $0 \leq x \leq 1$;}\\ &\text{otherwise.}
Notice how you can put inline math, opened and closed with $, inside the argument of a \text command. The rules for using the cases environment are the same as for matrices. Separate the lines with \\ and indicate the annotation with &.
CHAPTER
4 Your first article and presentation
4.1
The anatomy of an article To begin, we use the sample article intrart.tex (in the samples folder) to examine the anatomy of an article. Copy it over to the work folder or type it, and save it in the work folder as we discuss the parts of an article. Every LATEX article has two parts, the preamble and the body. The preamble of an article is everything from the first line of the source file down to the line \begin{document} For a schematic view of an article, see Figure 4.1. The preamble contains instructions affecting the entire document. The only required command in the preamble is the \documentclass command. There are other commands (such as the \usepackage commands) that must be placed in the preamble if they are used, but these commands do not have to be present in every document. Here is the preamble of the introductory sample article:
36
Chapter 4
Your first article and presentation
% Introductory sample article: intrart.tex \documentclass{amsart} \usepackage{amssymb,latexsym} \usepackage{graphicx} \newtheorem{theorem}{Theorem} \newtheorem{lemma}{Lemma} \newtheorem{definition}{Definition} \newtheorem{notation}{Notation} The preamble specifies the document class and then the LATEX enhancements, or packages, used in the article. The preamble can also specify additional commands that are used throughout the document, such as proclamation definitions, userdefined commands, and so on. intrart.tex specifies the amsart document class. This class defines the format used by the AMS journals—and many others—for articles. LATEX then loads two pack
\documentclass{...} \usepackage{...} ...
preamble
\begin{document}
top matter
\title{...} \author{...} \address{...} \date{...} \begin{abstract} ... \end{abstract} \maketitle
main matter
abstract
\section{...} \section{...}
\begin{thebibliography}{9} back matter ... \end{thebibliography} \end{document}
Figure 4.1: A schematic view of an article.
body
4.1
The anatomy of an article
37
ages, latexsym and amssymb, that provide the names of some mathematical symbols. Finally, LATEX loads the graphicx package, which we need because of the illustration. The preamble concludes with the proclamations. A proclamation is a theorem, lemma, definition, corollary, note, or other similar construct. The intrart.tex article defines four proclamations. The first of these, \newtheorem{theorem}{Theorem} defines the theorem environment, which then can be used in the body of the article, as explained in Section 4.2.3. The other three are similar. LATEX automatically numbers and formats proclamations. The article proper, called the body, is the content of the document environment— it is between the lines \begin{document} and \end{document} as illustrated in Figure 4.1. The body of an article is split into several parts, starting with the top matter containing title page information and the abstract. The top matter follows the line \begin{document} and concludes with the line \maketitle Here is the top matter of the introductory sample article: \title{A construction of completesimple\\ distributive lattices} \author{George~A. Menuhin} \address{Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \date{March 15, 2006} \begin{abstract} In this note, we prove that there exist \emph{completesimple distributive lattices,} that is, complete distributive lattices with only two complete congruences. \end{abstract} \maketitle
38
Chapter 4
Your first article and presentation
And here is the rest of the body of the introductory sample article with some commentary, exclusive of the bibliography: \section{Introduction}\label{S:intro} In this note, we prove the following result: \begin{theorem} There exists an infinite complete distributive lattice~$K$ with only the two trivial complete congruence relations. \end{theorem} \section{The $\Pi^{*}$ construction}\label{S:P*} The following construction is crucial in the proof of our Theorem (see Figure~\ref{Fi:products}): \begin{definition}\label{D:P*} Let $D_{i}$, for $i \in I$, be complete distributive lattices satisfying condition~\textup{(J)}. Their $\Pi^{*}$ product is defined as follows: \[ \Pi^{*} ( D_{i} \mid i \in I ) = \Pi ( D_{i}^{} \mid i \in I ) + 1; \] that is, $\Pi^{*} ( D_{i} \mid i \in I )$ is $\Pi ( D_{i}^{} \mid i \in I )$ with a new unit element. \end{definition} \begin{notation} If $i \in I$ and $d \in D_{i}^{}$, then \[ \langle \dots, 0, \dots, d, \dots, 0, \dots \rangle \] is the element of $\Pi^{*} ( D_{i} \mid i \in I )$ whose $i$th component is $d$ and all the other components are $0$. \end{notation} See also Ernest~T. Moynahan~\cite{eM57a}. Next we verify the following result:
4.1
The anatomy of an article
39
\begin{theorem}\label{T:P*} Let $D_{i}$, $i \in I$, be complete distributive lattices satisfying condition~\textup{(J)}. Let $\Theta$ be a complete congruence relation on $\Pi^{*} ( D_{i} \mid i \in I )$. If there exist $i \in I$ and $d \in D_{i}$ with $d < 1_{i}$ such that, for all $d \leq c < 1_{i}$, \begin{equation}\label{E:cong1} \langle \dots, d, \dots, 0, \dots \rangle \equiv \langle \dots, c, \dots, 0, \dots \rangle \pod{\Theta}, \end{equation} then $\Theta = \iota$. \end{theorem} We include an illustration, products.eps or products.pdf (in your samples folder). We copy them over to the work folder and load the graphicx package. We name the illustration in the figure environment. The illustration must be in the graphic image file formats EPS or PDF (see Chapter 13.1.2). We left the argument of the \caption command empty—it normally contains the name or a description of the figure. The illustration is centered with the \centering command (see Section 6.3). \begin{figure}[hbt] \centering\includegraphics{products} \caption{}\label{Fi:products} \end{figure} The figure environment floats, that is, LATEX decides where to place the typeset figure. We can influence LATEX’s choice (see Section 10.4.3). Then we place a proof in a proof environment. \begin{proof} Since \begin{equation}\label{E:cong2} \langle \dots, d, \dots, 0, \dots \rangle \equiv \langle \dots, c, \dots, 0, \dots \rangle \pod{\Theta}, \end{equation} and $\Theta$ is a complete congruence relation, it follows from condition~(J) that \begin{equation}\label{E:cong} \langle \dots, d, \dots, 0, \dots \rangle \equiv \bigvee ( \langle \dots, c, \dots, 0, \dots \rangle \mid d \leq c < 1 ) \pod{\Theta}.
40
Chapter 4
Your first article and presentation
\end{equation} Let $j \in I$, $j \neq i$, and let $a \in D_{j}^{}$. Meeting both sides of the congruence \eqref{E:cong2} with $\langle \dots, a, \dots, 0, \dots \rangle$, we obtain that \begin{equation}\label{E:comp} 0 = \langle \dots, a, \dots, 0, \dots \rangle \pod{\Theta}, \end{equation} Using the completeness of $\Theta$ and \eqref{E:comp}, we get: \[ 0 \equiv \bigvee ( \langle \dots, a, \dots, 0, \dots \rangle \mid a \in D_{j}^{} ) = 1 \pod{\Theta}, \] hence $\Theta = \iota$. \end{proof} At the end of the body, the bibliographic entries are typed between the lines \begin{thebibliography}{9} and \end{thebibliography} There are fewer than 10 references in this article, so we tell LATEX to make room for singledigit numbering by providing the argument 9 to the thebibliography environment. We use 99 if the number of references is between 10 and 99. The typeset bibliography is titled References. The bibliography of intrart.tex is structured as follows: \begin{thebibliography}{9} \bibitem{sF90} SooKey Foo, \emph{Lattice Constructions}, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. \bibitem{gM68} George~A. Menuhin,
4.1
The anatomy of an article
41
\emph{Universal algebra}. D.~Van Nostrand, Princeton, 1968. \bibitem{eM57} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \textbf{8} (1957), 455460. \bibitem{eM57a} Ernest~T. Moynahan, \emph{Ideals and congruence relations in lattices}. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K\"{o}zl. \textbf{9} (1957), 417434. \end{thebibliography} The body and the article end when the document environment is closed with \end{document}
4.1.1 The typeset sample article On the next two pages, you find the typeset intrart.tex, the introductory sample article.
42
Chapter 4
Your first article and presentation
A CONSTRUCTION OF COMPLETESIMPLE DISTRIBUTIVE LATTICES GEORGE A. MENUHIN Abstract. In this note, we prove that there exist completesimple distributive lattices, that is, complete distributive lattices with only two complete congruences.
1. Introduction In this note, we prove the following result: Theorem 1. There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations. 2. The Π∗ construction The following construction is crucial in the proof of our Theorem (see Figure 1): Definition 1. Let Di , for i ∈ I, be complete distributive lattices satisfying condition (J). Their Π∗ product is defined as follows: Π∗ (Di  i ∈ I) = Π(Di−  i ∈ I) + 1; that is, Π∗ (Di  i ∈ I) is Π(Di−  i ∈ I) with a new unit element. Notation 1. If i ∈ I and d ∈ Di− , then h. . . , 0, . . . , d, . . . , 0, . . . i ∗
is the element of Π (Di  i ∈ I) whose ith component is d and all the other components are 0. See also Ernest T. Moynahan [4]. Next we verify the following result: Theorem 2. Let Di , i ∈ I, be complete distributive lattices satisfying condition (J). Let Θ be a complete congruence relation on Π∗ (Di  i ∈ I). If there exist i ∈ I and d ∈ Di with d < 1i such that, for all d ≤ c < 1i , (1)
h. . . , d, . . . , 0, . . . i ≡ h. . . , c, . . . , 0, . . . i (Θ),
then Θ = ι. Proof. Since (2)
h. . . , d, . . . , 0, . . . i ≡ h. . . , c, . . . , 0, . . . i (Θ),
and Θ is a complete congruence relation, it follows from condition (J) that _ (3) h. . . , d, . . . , 0, . . . i ≡ (h. . . , c, . . . , 0, . . . i  d ≤ c < 1) (Θ). Date: March 15, 2006. 1
4.1
43
The anatomy of an article
2
GEORGE A. MENUHIN
Π(Di−  i ∈ I)
Π(Di−  i ∈ I)
Π(Di−  i ∈ I) + 1;
Π(Di  i ∈ I) Figure 1
Let j ∈ I, j 6= i, and let a ∈ Dj− . Meeting both sides of the congruence (2) with h. . . , a, . . . , 0, . . . i, we obtain that (4)
0 = h. . . , a, . . . , 0, . . . i (Θ),
Using the completeness of Θ and (4), we get: _ 0 ≡ (h. . . , a, . . . , 0, . . . i  a ∈ Dj− ) = 1 (Θ), hence Θ = ι.
References
[1] SooKey Foo, Lattice Constructions, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. [2] George A. Menuhin, Universal algebra. D. Van Nostrand, Princeton, 1968. [3] Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. [4] Ernest T. Moynahan, Ideals and congruence relations in lattices. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl. 9 (1957), 417–434. Computer Science Department, University of Winnebago, Winnebago, MN 53714
44
4.2
Chapter 4
Your first article and presentation
An article template In this section, you create an article template. To start a new article, open the template and start writing! Make a copy of intrart.tex and give the copy an appropriate name. I named mine gg.tex (it is in the samples folder). Remember the naming rule (page 4): The name should have no spaces, no special characters, and end with .tex. So the name my template.tex is not good, but MyTemplate.tex is fine.
4.2.1
Editing the top matter Edit the top matter to contain the relevant information, e.g., title and address, for your template. Here are some simple rules to follow.
Rule
Top matter 1. \thanks places an unmarked footnote at the bottom of the first page, for instance to acknowledge research support. If it is not needed, comment it out. 2. Separate the lines of your address with \\. Do not put a \\ at the end of the last line. 3. \date{\today} typesets today’s date. If you do not want any date to appear, comment out the \date command. For a specific date, such as March 15, 2006, type \date{March 15, 2006}. 4. The \title command is the only required command. The others are optional. Actually, if you do not give the \maketitle command, even the \title command is optional. On the other hand, if you do give the \maketitle command and omit the \title command, then you get the error message ! Undefined control sequence. <argument> \shorttitle l.27 \maketitle Now delete all the content of the article, leaving you with the skeleton. Here is the edited gg.tex (it is in your samples folder): % GG’s article template: gg.tex \documentclass{amsart} \usepackage{amssymb,latexsym} \usepackage{graphicx}
4.2
An article template
45
\newtheorem{theorem}{Theorem} \newtheorem{lemma}{Lemma} \newtheorem{definition}{Definition} \newtheorem{notation}{Notation} \begin{document} \title{Title!} \author{George Gr\"{a}tzer} \address{Department of Mathematics\\ University of Manitoba\\ Winnipeg, MB R3T 2N2\\ Canada} \date{\today} \begin{abstract} To come! \end{abstract} \maketitle \section{Introduction}\label{S:intro} \begin{thebibliography}{9} \end{thebibliography} \end{document} I also made a version for a joint article with another author: gg2.tex (see the samples folder). It adds the lines \author{Second author} \address{line1\\ line2\\ line3\\ line4} before the \date command. When I start writing an article, I open gg.tex or gg2.tex, save it under a new name, and edit the top matter. Here are two more rules about the top matter to keep in mind:
46
Rule
Chapter 4
Your first article and presentation
Top matter (continued) 5. If necessary, break the title into separate lines with \\. Do not put a \\ at the end of the last line. 6. Multiple authors get separate \author and \address commands.
4.2.2
Sectioning An article, as a rule, is divided into sections. To start the section Introduction, type \section{Introduction}\label{S:intro} after the \maketitle command. Introduction typesets as the title of the section. I use the convention that S: starts the label for a section, so the label is S:intro (or something similar). The section’s number is automatically assigned by LATEX. You can refer to this section number with \ref{S:intro}: In Section~\ref{S:intro}, we introduce The command \section* produces an unnumbered section. Sections have subsections, and subsections have subsubsections. The corresponding commands are \subsection
\subsubsection
Their unnumbered variants are \subsection*
4.2.3
\subsubsection*
Invoking proclamations In the preamble of the article intrart.tex, we defined the theorem, lemma, definition, and notation proclamations. These proclamations define environments. For example, you type a theorem within a theorem environment. The body of the theorem, that is, the part of the source file that produces the theorem, is typed between the lines \begin{theorem}\label{T:xxx } and \end{theorem} where T:xxx is the label for the theorem. You should replace xxx with a label that is somewhat descriptive of the contents of your theorem. LATEX automatically assigns a number to the theorem, and the theorem can be referenced by using a command of the form \ref{T:xxx }.
4.2
An article template
47
4.2.4 Inserting references The works to be listed are placed in the bibliography. Below are typical entries for the most frequently used types of references, an article in a journal, a book, a Ph.D. thesis, and a technical report. For more examples, see the bibliographic template file, bibl.tpl, in the samples folder. \begin{thebibliography}{9} \bibitem{sF90} SooKey Foo, \emph{Lattice Constructions}, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. \bibitem{gM68} George~A. Menuhin, \emph{Universal algebra}. D.~Van Nostrand, Princeton, 1968. \bibitem{eM57} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \textbf{8} (1957), 455460. \bibitem{eM57a} Ernest~T. Moynahan, \emph{Ideals and congruence relations in lattices.} II, Magyar Tud. Akad. Mat. Fiz. Oszt. K\"{o}zl. \textbf{9} (1957), 417434. \end{thebibliography} Each item listed in the bibliography can be referenced in the body of the article. You reference with the \cite command. The argument is the argument of the \bibitem command. So to reference Menuhin’s article, type \cite{gM68} which typesets as [2] since Menuhin’s article is the second in the list. So this result was first published in [2] is typed as this result was first published in~\cite{gM68}
48
Chapter 4
Your first article and presentation
How you write each label is up to you, subject only to the rule in Section 10.4.2, provided the labels are unique. I use the convention that the label for a \bibitem consists of the initials of the author and the year of publication. For example, a publication by Andrew B. Reich in 1987 would have the label aR87 (a second publication by that author from that year would be aR87a). For joint publications, the label consists of the initials of the authors and the year of publication. For example, a publication by John Bradford and Andrew B. Reich in 1987 would have the label BR87. You have to arrange the references in your document’s thebibliography environment in the order you wish to see them. LATEX only takes care of the numbering and the citations in the text.
4.3
On using LATEX Now that you are ready to type your first article, we give you some pointers on using LATEX.
4.3.1
LATEX error messages You probably make a number of mistakes in your first article. These mistakes fall into the following categories: 1. Typographical errors, which LATEX blindly typesets 2. Errors in mathematical formulas or in the formatting of the text 3. Errors in your instructions to LATEX, that is, in commands and environments Typographical errors can be corrected by viewing and spell checking the typeset article, finding the errors, and then editing the source file. Mistakes in the second and third categories probably trigger errors during the typesetting process, such as the math errors in Section 3.2. Some of these mistakes may have to be corrected before your article can be completely typeset. We now look at some examples of the third class of errors by deliberately introducing a number of mistakes into the source file of the article intrart.tex (in your samples folder, source file on pages 35–41, and shown typeset on pages 42–43), and examining the error messages that occur. When LATEX displays a ? prompt, you can either try to continue typesetting the document by pressing Return, or type x to stop typesetting immediately. See Section D.4 for other options. Experiment 1 In intrart.tex, go to line 20 by using your editor’s Go to Line command and remove the closing brace so that it reads \begin{abstract
4.3
On using LATEX
49
When you typeset intrart.tex, LATEX reports a problem: Runaway argument? {abstract In this note, we prove that there exist \emph {completesim\ETC. ! Paragraph ended before \begin was complete. \par l.26 Line 26 of the file is the line after \end{abstract}. The error message informs you that the name of the environment that ends before line 26 is not completed before the end of the paragraph. Press Return to tell LATEX to typeset the remainder of the article, leaving out the abstract. Runaway argument? is an error message that comes up often. It usually means that the argument of a command is either longer than expected or it contains material that the argument cannot accept. Most often a closing brace solves the problem, as in the experiment. Experiment 2 Now correct line 20, then go to line 25 and change it from \end{abstract} to \end{abstrac} and typeset the article again. LATEX informs you of another error: ! LaTeX Error: \begin{abstract} on input line 20 ended by \end{abstrac}. l.25 \end{abstrac} This is perfect. LATEX correctly analyzes the problem and tells you where to make the change. You may continue typesetting the article by pressing Return. LATEX then gives you the message: ! Missing } inserted. } l.25 \end{abstrac} The missing } inserted is the “special brace” \end{abstract} (more about this in Section 5.3.2). Press Return. LATEX recovers from this error and the article is typeset correctly.
50
Chapter 4
Your first article and presentation
Experiment 3 Instead of correcting the error in line 25, comment it out with % \end{abstrac} and also comment out the four lines of the figure environment using block comment. Introduce an additional error in line 96. This line reads and $\Theta$ is a complete congruence relation, Change \Theta to \Teta: and $\Teta$ is a complete congruence relation, Now, when you typeset the article, LATEX reports ! Undefined control sequence. \Teta l.96 and $\Teta $ is a complete congruence relation, Pressing Return results in the message ! LaTeX Error: \begin{abstract} on input line 20 ended by \end{document}. Type ...
H
for immediate help.
l.150 \end{document} These two mistakes are easy to identify: \Teta is a misspelling of \Theta, and since \end{abstract} is missing, LATEX is trying to match \begin{abstract} with \end{document} Now undo the changes you made to lines 25 and 96. Uncomment the figure. Experiment 4 In line 42, delete the closing brace of the \label command: \begin{definition}\label{D:P* This results in a message for line 54, the blank line following the paragraph, that ! Paragraph ended before \label was complete.
4.3
On using LATEX
51
This is easy to understand. You cannot begin a new paragraph within the argument of a \label command. Undo the change to line 42. Experiment 5 Add a blank line following line 58: \langle \dots, 0, \dots, d, \dots, 0, \dots \rangle This change results in the message ! Missing $ inserted. $ l.59 There can be no blank lines within a displayed math environment. LATEX catches the mistake, but the message itself is misleading. Pressing Return does not help; you cannot recover from the error. Delete the blank line. Experiment 6 Add a $ somewhere in line 58 (such errors often occur when cutting and pasting formulas): \langle $\dots, 0, \dots, d, \dots, 0, \dots \rangle You get the message: ! Display math should end with $$. \protect l.58 \langle $\dots , 0, \dots, d, \dots, 0, \dots \rangle Maybe this could be more to the point? Error messages from LATEX are not always helpful, but there is always some information that can be gleaned from them. Try to identify the structure, that is, the command or environment, that causes the error—read the section of this book that describes that command or environment. This should help you correct the error. Keep in mind that the error could be quite far from the line LATEX indicates, but it is always on or before that line in the source file. If you have difficulty isolating a problem, block comment all but the paragraph you suspect might have problems. If necessary, split a large paragraph into smaller pieces.
Tip
Typeset often.
52
Chapter 4
Your first article and presentation
To some extent, you can avoid having to isolate problems by following this tip. For instance, if I were to typeset First Steps into LATEX [29], with the closing brace of the first \caption command on line 480 of the source file missing, I would get the error message ! Text line contains an invalid character. l.1227 ...pletely irreducible^^? where the reference is to line 1227, about 700 lines removed from the actual error. However, if the only thing I did before typesetting was to insert that figure with its incorrect caption command, at least I would know where to look for errors. If you make a dozen corrections and then typeset, you may not know where to start.
4.3.2
Logical and visual design The typeset version of intrart.tex (pp. 42–43) looks impressive. To produce such articles, you need to understand that there are two aspects to article design: visual and logical. As an example, let us look at a theorem from intrart.tex (see the typeset form of the theorem on page 43). You tell LATEX that you want to state a theorem by using a theorem environment: \begin{theorem}\label{T:P*} Let $D_{i}$, $i \in I$, be complete distributive lattices satisfying condition~\textup{(J)}. Let $\Theta$ be a complete congruence relation on $\Pi^{*} ( D_{i} \mid i \in I )$. If there exist $i \in I$ and $d \in D_{i}$ with $d < 1_{i}$ such that, for all $d \leq c < 1_{i}$, \begin{equation}\label{E:cong1} \langle \dots, d, \dots, 0, \dots \rangle \equiv \langle \dots, c, \dots, 0, \dots \rangle \pod{\Theta}, \end{equation} then $\Theta = \iota$. \end{theorem} The logical part of the design is choosing to define a theorem by placing material inside a theorem environment. For the visual design, LATEX makes hundreds of decisions. Could you have specified all of the spacing, font size changes, centering, numbering, and so on? Maybe, but would you want to? And would you want to repeat that process for every theorem in your document? Even if you did, you would have spent a great deal of time and energy on the visual design of the theorem rather than on the logical design of your article. The idea
4.4
Converting an article to a presentation
53
behind LATEX is that you should concentrate on what you have to say and let LATEX take care of the visual design. This approach allows you to easily alter the visual design by changing the document class (or its options, see Sections 11.5, 12.1.2, and 18.1). Section 11.1 provides some examples. If you code the visual design into the article—hard coding it, as a programmer would say—such changes are much harder to accomplish, for you and for the journal publishing the article.
4.4
Converting an article to a presentation To produce a document in LATEX for use as a presentation, you have to output it as a PDF file. You make your presentation using a PDF viewer such as Adobe Reader or print the pages of the PDF file on transparencies and use a projector. So a presentation is a PDF file. To display the presentation, connect your computer to a projector. Open the PDF file in Adobe Reader, put it in full screen mode. Then project the presentation a page at a time by pressing the space bar or the forward and back arrow keys. In LATEX, you use a presentation package—really, a document class—to prepare such a PDF file. We deal with presentations in detail in Chapter 14, but as a quick introduction, we convert intrart.tex into a presentation. For the conversion, we use the presentation package FoilTEX, while in Chapter 14 we discuss the beamer package. To use the FoilTEX package, we have to learn only one new command, \foilhead. Open intrart.tex, save it as intropres.tex in the work folder. We introduce some changes to the document to prepare it for the conversion. Once you are satisfied with the changes made, the tex file created for FoilTEX is typeset so as to produce the PDF file. For WinEdt, click on the PDF TeXify icon. For TeXShop, just click on Typeset. For other TEX installations, check your user manual on how to create a PDF output.
4.4.1 Preliminary changes Make the following changes in the preamble, top matter, and abstract. 1. Change the first line to % Introductory presentation:intropres.tex 2. Change the documentclass to foils. 3. Add the line \usepackage{amsmath} after the documentclass line. We have to do this because FoilTEX does not automatically load the AMS math package.
54
Chapter 4
Your first article and presentation
4. Delete the definitions of theorem, lemma, and definition. FoilTEX redefines these. 5. Copy the address into the \author command: \author{George~A. Menuhin\\ Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} and delete the \address command. This may seem strange, but it is necessary because FoilTEX is based on a legacy document class that does not have an \address command (see Chapter 12). 6. Move the abstract after the \maketitle command, as was customary in legacy document classes. 7. Add the [scale=2] option to the \includegraphics command, so the command becomes \centering\includegraphics[scale=2]{products} FoilTEX uses fonts in 20 point size, twice the usual size. So it is appropriate that we scale up the illustration to 200%. So the new version is % Introductory presentation:intropres.tex \documentclass{foils} \usepackage{amsmath} \usepackage{amssymb,latexsym} \usepackage{graphicx} \begin{document} \title{A construction of completesimple\\ distributive lattices} \author{George~A. Menuhin\\ Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \date{March 15, 2006} \maketitle \begin{abstract} In this presentation, we prove that there exist
4.4
Converting an article to a presentation
55
\emph{completesimple distributive lattices,} that is, complete distributive lattices with only two complete congruences. \end{abstract} Declarations in the body In your LATEX editor, perform four search and replace operations in the body of the article. Change all of the following: 1. {theorem} to {Theorem} 2. {lemma} to {Lemma} 3. {definition} to {Definition} 4. {proof} to {Proof} FoilTEX defines and uses the capitalized versions. Sectioning Comment out all the \section commands. FoilTEX uses the \foilhead command to break the material into pages and also as a substitute for sectioning.
4.4.2 Making the pages We cut the presentation into pages (transparencies or foils) by inserting as many page breaking commands of FoilTEX, \foilhead{}, as seems appropriate. The argument of the command becomes the “title” for the page. If the argument is empty, the page has no title. Add the \foilhead{The result} command after the abstract. This ends the title page and adds the title The result to the next page. See the intropres.tex document for all the other \foilhead{} commands we have added.
4.4.3 Fine tuning We have eliminated the equation numbering, because it would make the equations too wide. Also, in a presentation, references to another page are not recommended. We made some additional changes to accommodate that we have fewer characters per line. Compare the documents intrart.tex and intropres.tex to see all the changes. Note how in the PDF file the fonts are automatically changed to sans serif, because sans serif text is easier to read when projected. The font size is 20pt, twice the size of the font in the article.
56
Chapter 4
Your first article and presentation
Of course, intropres.tex is not the most elegant presentation. But I hope it helps you to make your first presentation. On pages 57 and 58, we show the first two pages of this presentation. There are, of course, a number of useful commands in FoilTEX in addition to the one we used, \foilhead. We did not even do justice to this one command. It has an optional argument to enlarge or shrink the space between the header and the body of the foil. So \foilhead[.5in]{A diagram} shrinks that space by half an inch. This is especially useful with large diagrams. For numerous other features of FoilTEX, see the user manual [33].
4.4
Converting an article to a presentation
57
58
Chapter 4
Your first article and presentation
CHAPTER
5 Typing text
In Chapter 2, we briefly discussed how to type text in a document. Now we take up this topic more fully. This chapter starts with a discussion of the keyboard in Section 5.1 and continues with the rules for spaces in Section 5.2. We cover a very important topic that must precede any indepth discussion of LATEX, how to control LATEX with commands and environments, in Section 5.3. A document may contain symbols that cannot be found on your keyboard. In Section 5.4, we show how to get these symbols in our typeset documents by using commands. Some other characters are defined by LATEX as command characters. For example, the % character plays a special role in the source document. In Section 5.5.1, you will see how % is used to comment out lines. In Section 5.5.2, we introduce the command for footnotes. In Section 5.6, we discuss the commands (and environments) for changing fonts, their shapes and sizes. In Section 5.7, you learn about lines, paragraphs, and pages. The judicious use of horizontal and vertical spacing is an important part of document
62
Chapter 5
Typing text
formatting, and also the topic of Section 5.8. In Section 5.9, you learn how to typeset text in a “box”, which behaves as if it were a single large character. To help the discussion along, we shall use the terms text mode and math mode to distinguish between typesetting text and math.
5.1
The keyboard Most of the keys on your computer’s keyboard produce characters, while others are function or modifier keys.
5.1.1
Basic keys The basic keys are grouped as follows: Letters The 52 letter keys: a
b
c ... z
A
B
C ... Z
Digits The ten digits: 1
2 ... 9
0
Oldstyle digits are available with the \oldstylenums command. The next line shows the default digits followed by the old style digits: 1234567890
typed as 1234567890 \quad \oldstylenums{1234567890} Punctuation There are nine punctuation marks: ,
;
.
?
!
:
‘
’

The first six are the usual punctuation marks. The ‘ is the left single quote— also known as the grave accent—while ’ doubles as the right single quote and apostrophe (see Section 5.4.1). The  key is the dash or hyphen (see Sections 5.4.2 and 5.4.9). Parentheses There are four: (
)
[
]
( and ) are parentheses; [ and ] are called (square) brackets.
5.1
63
The keyboard Math symbols Seven math symbols correspond to keys. The math symbols are: ∗
+
=
−
/
The last four characters have a role also in text mode: The minus sign − corresponds to the hyphen key,  (see Section 7.4.1). The math symbols < and > correspond to the keys < and >; use them only in math mode. Note that there is also a version of colon (:) for math formulas (see Sections 8.1 and B.2). Space keys Pressing the space bar gives the space character. Pressing the tab key gives the tab character. When typesetting the source file, LATEX does not distinguish between these two. Pressing the Return key gives the endofline character. These keys produce invisible characters that are normally not displayed on your monitor by the text editor. Different computer systems have different endofline characters, which may cause some problems when transferring files from one system to another. A good text editor translates endofline characters automatically or on demand. Section 5.2.1 explains how LATEX handles the invisible characters. When explaining some rules of LATEX, sometimes it is important to show if a space is required. In such cases, I use the symbol to indicate a space, for instance, \in ut and \ . The tilde ~ produces a nonbreakable space or tie (see Section 5.4.3 and the footnote on page 11).
5.1.2 Special keys There are 13 special keys on the keyboard: #
$
%
&
~
_
^
\
{
}
@
"

They are mostly used to give instructions to LATEX and some are used in math mode (see Chapter 7), and some in B IBTEX (see Chapter 16). See Section 5.4.4 on how to print these characters in text. Only @ requires no special command, type @ to print @.
5.1.3 Prohibited keys Keys other than those discussed in Sections 5.1.1 and 5.1.2 are prohibited! Specifically, do not use the computer’s modifier keys—Control, Alt, Escape, and others—to produce special characters, such as accented characters. LATEX will either reject or misunderstand them.
64
Chapter 5
Typing text
Prohibited characters may not cause problems in some newer LATEX implementations. They may just print Ł if your source file has Ł, and ignore the invisible invalid characters. However, for portability reasons, you should avoid using prohibited characters. The babel package provides support for using some modifier keys (see Appendix G).
Tip
If there is a prohibited character in your document, you may receive an error message such as ! Text line contains an invalid character. l.222 completely irreducible^^? ^^? Delete and retype the offending word or line until the error goes away.
5.2
Words, sentences, and paragraphs Text consists of words, sentences, and paragraphs. In text, words are separated by one or more spaces, which may include a single endofline character (see the rule, Spacing in text), or by parentheses and punctuation marks. A group of words terminated by a period, exclamation point, or question mark forms a sentence (not all periods terminate a sentence, see the discussion in Section 5.2.2). A group of sentences terminated by one or more blank lines constitutes a paragraph.
5.2.1
Spacing rules Here are the most important LATEX rules about spaces in text in the source file.
Rule
Spacing in text 1. Two or more spaces in text are the same as one. 2. A tab or endofline character is the same as a space. 3. A blank line, that is, two endofline characters separated only by spaces and tabs, indicates the end of a paragraph. The \par command is equivalent. 4. Spaces at the beginning of a line are ignored. Rules 1 and 2 make cutting and pasting text less errorprone. In your source file, you do not have to worry about the line length or the number of spaces separating words
5.2
Words, sentences, and paragraphs
65
or sentences, as long as there is at least one space or endofline character separating any two words. Thus You do not have to worry about the number of spaces separating words, as long as there is at least one space or endofline character separating any two words. produces the same typeset text as You do not have to worry about the number of spaces separating words, as long as there is at least one space or endofline character separating any two words. However, the number of as long
spaces separating words,
and the number of , as long
spaces separating words
produce different results:
the number of spaces separating words, as long the number of spaces separating words , as long Notice the space between “words” and the comma in the second line. That space was produced by the endofline character in accordance with Rule 2. It is very important to maintain the readability of your source file. LATEX may not care about the number of spaces or line length, but you, your coauthor, or your editor might. Rule 3 contradicts Rules 1 and 2, consider it an exception. Sometimes—especially when defining commands and environments (see Sections 15.1 and 15.2)—it is more convenient to indicate the end of a paragraph with \par. When emailing a source file, email clients tend to break longer lines into shorter ones. Because of the space rules, this does not effect the typeset version most of the time.
66
Chapter 5
5.2.2
Typing text
Periods LATEX places a certain size space between words—the interword space—and a somewhat larger space between sentences—the intersentence space. To know which space to use, LATEX must decide whether or not a period indicates the end of a sentence.
Rule 1
Period To LATEX, a period after a capital letter, for instance, A. or caT., signifies an abbreviation or an initial. Generally, every other period signifies the end of a sentence. This rule works most of the time. When it fails—for instance, twice with e.g.—you need to specify the type of space you want, using the following two rules.
Rule 2
Period If an abbreviation does not end with a capital letter, for instance, etc., and it is not the last word in the sentence, then follow the period by an interword space (\ ) or a tie (~), if appropriate (see Section 5.4.3). Recall that \ provides an interword space. The result was first published, in a first approximation, in the Combin.\ Journal. The result was first published, in a first approximation, in the Combin. Journal. prints as The result was first published, in a first approximation, in the Combin. Journal. The result was first published, in a first approximation, in the Combin. Journal. Notice that Combin. in the first line is followed by a regular interword space. The intersentence space following Combin. in the second line is a little wider. A tie (or nonbreakable space)—see also Section 5.4.3—is more appropriate than \ in phrases such as Prof. Smith, typed as Prof.~Smith, and pp. 271–292, typed as pp.~271292.
Tip
The thebibliography environment handles periods properly. You do not have to mark periods for abbreviations (in the form .\) in the name of a journal, so Acta Math. Acad. Sci. Hungar. is correct.
5.3
Commanding LATEX
Rule 3
67
Period If a capital letter is followed by a period and is at the end of a sentence, precede the period with \@. For example, (1) follows from condition~H\@. We can proceed\\ (1) follows from condition~H. We can proceed prints: (1) follows from condition H. We can proceed (1) follows from condition H. We can proceed Notice that there is not enough space after H. in the second line. Most typographers agree on the following rule (see, e.g., Robert Bringhurst’s The Elements of Typographic Style [8], p. 30):
Rule 4
Period Add no space or a thin space (\,) within strings of initials and be consistent. So W.H. Lampstone with no space or W. H. Lampstone with thin space is preferred over W. H. Lampstone. My personal choice is W. H. Lampstone with thin space. To make all intersentence spaces equal to the interword space—as required in French typography—you can use the command \frenchspacing To switch back to using spaces of different sizes, give the command \nonfrenchspacing
5.3
Commanding LATEX How do you command LATEX to do something special for you, such as starting a new line, changing emphasis, or displaying the next theorem? You use commands and special pairs of commands called environments, both briefly introduced at the start of Chapter 2. Most, but not all, commands have arguments, which are usually fairly brief. Environments have contents, the text between the \begin and \end commands. The contents of an environment can be several paragraphs long.
68
5.3.1
Chapter 5
Typing text
Commands and environments The \emph{text } command instructs LATEX to emphasize its argument, text. The \& command has no argument. It instructs LATEX to typeset & (see Section 5.4.4). The flushright environment instructs LATEX to right justify the content, the text between the two commands \begin{flushright} \end{flushright} The content of the document environment is the body of the article (see Section 4.1) and the content of the abstract environment is the abstract.
Rule
Environments An environment starts with the command \begin{name } and ends with \end{name } Between these two lines is the content of the environment, affected by the definition of the environment.
Rule
Commands A LATEX command starts with a backslash, \, and is followed by the command name. The name of a command is either a single nonalphabetic character other than a tab or endofline character or a string of letters, that is, one or more letters. So # and ’ are valid command names. The corresponding commands \# and \’ are used in Sections 5.4.4 and 5.4.7, respectively. input and date are also valid command names. However, input3, in#ut, and in ut are not valid names because 3, #, and should not occur in a multicharacter command name. Note that is a command name, the command \ produces a blank. LATEX has a few commands, for instance, $ (see Section 7.1) that do not follow this naming scheme, that is, they are not of the form \name. See also Section 15.1.8 for special commands with special termination rules.
5.3
Commanding LATEX
Rule
69
Command termination LATEX finds the end of a command name as follows: If the first character of the name is not a letter, the name is the first character. If the first character of the name is a letter, the command name is terminated by the first nonletter. If the command name is a string of letters, and is terminated by a space, then LATEX discards all spaces following the command name. While input3 is an invalid name, \input3 is not an incorrect command. It is the \input command followed by the character 3, which is either part of the text following the command or the argument of the command. LATEX also allows some command names to be modified with *. Such commands are referred to as *ed commands. Many commands have *ed variants. \hspace* is an oftenused *ed command (see Section 5.8.1).
Rule
Command and environment names Command and environment names are case sensitive. \ShowLabels is not the same as \showlabels.
Rule
Arguments Arguments are enclosed in braces, { }. Optional arguments are enclosed in brackets, [ ]. Commands may have arguments, typed in braces immediately after the command. The argument(s) are used in processing the command. Accents provide very simple examples. For instance, \’{o}—which produces ´o—consists of the command \’ and the argument o (see Section 5.4.7). In \bibliography{article1}, the command is \bibliography and the argument is article1 (see Section 16.2.2). Sometimes, if the argument is a single character, the braces can be dropped: \’o also typesets as ´ o. Some environments also have arguments. For example, the alignat environment (see Section 9.5.4) is delimited by the commands \begin{alignat}{2} and \end{alignat}
70
Chapter 5
Typing text
The argument, 2, is the number of columns—it could be any number 1, 2, . . . A command or environment may have more than one argument. The \frac command (see Section 7.4.1) has two, $\frac{1}{2}$ typesets as 12 . The userdefined command \con has three (see Section 15.1.2). Some commands and environments have one or more optional arguments, that is, arguments that may or may not be present. The \sqrt command (see Section 7.4.5) √ has an optional argument for specifying roots other than the square root. To get 3 25, type \sqrt[3]{25}. The \documentclass command has an argument, the name of a document class, and an optional argument, a list of options (see Section 10.2), for instance, \documentclass[12pt,draft,leqno]{amsart}
Tip
If you get an error when using a command, check that: 1. The command is spelled correctly, including the use of uppercase and lowercase letters. 2. You have specified all required arguments in braces. 3. Any optional argument is in brackets, not braces or parentheses. 4. The command is properly terminated. 5. The package providing the command is loaded with the \usepackage command.
Most errors in the use of commands are caused by breaking the termination rule. We can illustrate some of these errors with the \today command, which produces today’s date. You have already seen this command in Section 2.4 (see also Section 5.4.8). The correct usage is \today\ is the day or \today{} is the day which both typeset as July 19, 2006 is the day In the first case, \today was terminated by \ , the command that produces an interword space. In the second case, it was terminated by the empty group { }.
5.3
Commanding LATEX
71
If there is no space after the \today command, as in \todayis the day you get the error message ! Undefined control sequence. l.3 \todayis the day LATEX thinks that \todayis is the command, and, of course, does not recognize it. If you type one or more spaces after \today: \today
is the day
LATEX interprets the two spaces as a single space by the first space rule (see page 64), and uses that one space to delimit \today from the text that follows it. So LATEX produces July 19, 2006is the day Section 15.1.8 discusses how best to avoid such errors.
Tip
If a command—or environment—can have an optional argument and none is given, and the text following the command starts with [, then type this as {[}. This may happen, for instance, with the command \item (see page 123). To get an example for an environment, see Section 9.6.1 for subsidiary math environments and page 229 for the alignment options. See what happens if no option is given but the math starts, say, with [x].
5.3.2 Scope A command issued inside a pair of braces { } has no effect beyond the right brace, except for the seldom occurring global commands (see Section 5.3.3). You can have any number of pairs of braces: { ... { ... { ... } ... } ... } The innermost pair containing a command is the scope of that command. The command has no effect outside its scope. We can illustrate this concept using the \bfseries command that switches the font to boldface: {some text \bfseries bold text} no more bold
72
Chapter 5
Typing text
typesets as some text bold text no more bold The commands \begin{name } and \end{name } bracketing an environment act also as a pair of braces. In particular, $, \[, and \] are special braces.
Rule
Braces 1. Braces must be balanced: An opening brace has to be closed, and a closing brace must have a matching opening brace. 2. Pairs of braces cannot overlap. Violating the first brace rule generates warnings and error messages. If there is one more opening brace than closing brace, the document typesets, but you get a warning: (\end occurred inside a group at level 1) For two or more unmatched opening braces, you are warned that \end occurred inside a group at level 2, and so on. There is a tendency to disregard such warnings since your article is already typeset and the error may be difficult to find. However, such errors may have strange consequences. At one point in the writing of my second LATEX book, there were two extra opening braces in Chapter 2. As a result, the title of Chapter 7 was placed on a page by itself! So it is best not to disregard such warnings. If you have one unmatched closing brace, you get an error message such as ! Too many }’s If special braces, say, \begin{name} and \end{name}, do not balance, you get an error message such as those discussed in Section 4.3.1: ! LaTeX Error: \begin{name} on input line 21 ended by \end{document}. or ! LaTeX Error: \begin{document} ended by \end{name}. To illustrate the second rule, here are two simple examples of overlapping braces. Example 1 {\bfseries some text \begin{lemma}
5.3
Commanding LATEX
73
more text} final text \end{lemma}
Example 2 {some \bfseries text, then math: $\sqrt{2} }, \sqrt{3}$ In Example 1, the scope of \bfseries overlaps the braces \begin{lemma} and \end{lemma}, whereas in Example 2, the scope of \bfseries overlaps the special braces $ and $. Example 1 is easy to correct: {\bfseries some text} \begin{lemma} {\bfseries more text} final text \end{lemma} Example 2 may be corrected as follows: {some \bfseries text, then math:} $\sqrt{2}, \sqrt{3}$ √ Actually, 2 does not even have a bold version (see Section 8.3.3). If the braces do overlap and they are of the same kind, LATEX simply misunderstands the instructions. The closing brace of the first pair is regarded as the closing brace of the second pair, an error that may be difficult to detect. LATEX can help if special braces overlap. Typesetting Example 1 gives the error message ! Extra }, or forgotten \endgroup. l.7 more text } final text
5.3.3 Types of commands It may be useful at this point to note that commands can be of various types. Some commands have arguments, and some do not. Some commands effect change only in their arguments, while some commands declare a change. For instance, \textbf{This is bold} typesets the phrase This is bold in bold type: This is bold and has no effect on the text following the argument of the command. On the other hand, the command \bfseries declares that the text that follows should be bold. This command has no argument. I call a command that declares change a command declaration. So \bfseries is a command declaration, while \textbf is not. As a rule, command declarations are commands without arguments. Commands with arguments are called long if their argument(s) can contain a blank line or a \par command; otherwise they are short. For example, \textbf is a short
74
Chapter 5
Typing text
command. So are all the top matter commands discussed in Section 11.2. The \parbox command, discussed in Section 5.9.4, is long. Finally, as discussed in Section 5.3.2, the effect of a command remains within its scope. This is true only of local commands. There are also some global commands, such as the \setcounter command described in Section 15.5.1. Fragile commands As a rule, LATEX reads a paragraph of the source file, typesets it, and then goes on to the next paragraph (see Section D.5). Some information from the source file, however, is separately stored for later use. Examples include the title of an article, which is reused as a running head (Section 11.2.1); titles of parts, sections, subsections, and other sectioning commands, which are used in the table of contents (Sections 18.2 and 10.4.1); footnotes (Section 5.5.2); table and figure captions (Section 10.4.3), which are used in lists of tables and figures (Section 10.4.3); and index entries (Chapter 17). These are movable arguments, and certain commands embedded in them must be protected from damage while being moved. LATEX commands that need such protection are called fragile. The inline math delimiter commands (introduced on page 18) \( and \) are fragile, while $ is not. In a movable argument, fragile commands must be protected with a \protect command. Thus The function \( f(x^{2}) \) is not an appropriate section title, but The function \protect \( f(x^{2}) \protect \) is. So is The function $f(x^{2})$ To be on the safe side, you should protect every command that might cause problems in a movable argument. Section 18.2 shows an example of what happens if a fragile command is not protected. Alternatively, use commands declared with \DeclareRobustCommand This command works the same way as \newcommand but the command defined is robust, that is, not fragile.
5.4
Symbols not on the keyboard A typeset document may contain symbols that cannot be typed. Some of these symbols may even be available on the keyboard but you are prohibited from using them (see
5.4
Symbols not on the keyboard
75
Section 5.1.3). In this section, we discuss the commands that typeset some of these symbols in text.
5.4.1 Quotation marks To produce single and double quotes, as in ‘subdirectly irreducible’ and “subdirectly irreducible” type ‘subdirectly irreducible’ and ‘‘subdirectly irreducible’’ Here, ‘ is the left single quote and ’ is the right single quote. Note that the double quote is obtained by pressing the single quote key twice, and not by using the double quote key. If you need single and double quotes together, as in “She replied, ‘No.’ ”, separate them with \, (which provides a thin horizontal space): ‘‘She replied, ‘No.’\,’’
5.4.2 Dashes Dashes come in three lengths. The shortest dash, called a hyphen, is used to connect words: MeanValue Theorem This phrase is typed with a single dash: MeanValue Theorem A mediumsized dash, called an en dash, is typed as  and is used For number ranges; for instance, the phrase see pages 23–45, is typed as see pages~2345 Note: ~ is a nonbreakable space or tie (see Section 5.4.3). In place of a hyphen in a compound adjective when one of the elements of the adjective is an open compound (such as New York) or hyphenated (such as nonEnglish). For instance, the phrase Jonathan Schmidt–Freid adjoint, is typed as Jonathan SchmidtFreid adjoint A long dash—called an em dash—is used to mark a change in thought or to add emphasis to a parenthetical clause, as in this sentence. The two em dashes in the last sentence are typed as follows:
76
Chapter 5
Typing text
A long dashcalled an \emph{em dash}is used In math mode, a single dash is typeset as the minus sign − (a binary operation) with some spacing on both sides, as in 15 − 3 or the “negative” as in −3 (see Sections 5.1.1 and 7.4.1). Note that there is no space before or after an en dash or em dash.
5.4.3
Ties or nonbreakable spaces A tie or nonbreakable space (sometimes called a blue space) is an interword space that cannot be broken across lines. For instance, when referencing P. Neukomm in an article, you do not want the initial P. at the end of a line and the surname Neukomm at the beginning of the next line. To prevent such an occurrence, you should type P.~Neukomm. If your keyboard does not have ~, use the \nobreakspace command instead, and type P.\nobreakspace Neukomm. The following examples show some typical uses: Theorem~\ref{T:main} in Section~\ref{S:intro} Donald~E. Knuth assume that $f(x)$ is (a)~continuous, (b)~bounded the lattice~$L$. Sections~\ref{S:modular} and~\ref{S:distributive} In~$L$, we find Of course, if you add too many ties, as in Peter~G.~Neukomm% Incorrect! LATEX may send you a line too wide error message (see Section 5.7.1). The tie (~) absorbs spaces, so typing P. ~ Neukomm works just as well. This feature is convenient when you add a tie during editing.
5.4.4
Special characters The characters corresponding to nine of the 13 special keys (see Section 5.1.2) are produced by typing a backslash (\) and then the key, as shown in Table 5.1. If for some reason you want to typeset a backslash in your document, type the command \textbackslash, which typesets as \. You might think that you could get a typewriter style backslash by utilizing the \texttt command introduced in Section 2.4
5.4
77
Symbols not on the keyboard \texttt{\textbackslash}
but this is not the case, \textbackslash and \texttt{\textbackslash} produce the same symbol, \, which is different from the typewriter style backslash: \. Look at them side by side: \ \. For a typewriter style backslash you can use the \bsl command introduced in Section 15.1.1 or the \texttt{\symbol{92}} command introduced later in this section. The  key is seldom used in text. If you need to typeset the math symbol  in text, type \textbar. Note that in text, * typesets as *, whereas in a formula it typesets centered as ∗. To typeset a centered star in text, use the command \textasteriskcentered. @ typesets as @. Finally, the " key should never be used in text. See Section 5.4.1 for the proper way to typeset double quotes. Nevertheless, sometimes " may be used to typeset ", as in the computer code segment print("Hello!"). In B IBTEX and MakeIndex, " has special meanings (see Chapters 16 and 17).
Tip
Be careful when typing \{ and \} to typeset the braces { }. Typing a brace without its backslash results in unbalanced braces, in violation of the first brace rule in Section 5.3.2. We illustrated in Section 5.3.2 some consequences of unbalanced braces. You may avoid some of these problems by introducing userdefined commands, as introduced in Section 15.3. You can also produce special characters with the \symbol command: \symbol{94} typesets as ˆ \symbol{126} typesets as ˜ Name
Type
Typeset
Ampersand Caret Dollar Sign Left Brace Right Brace Underscore (or Lowline) Octothorp Percent Tilde
\& \^{} \$ \{ \} \_ \# \% \~{}
& ˆ $ { } # % ˜
Table 5.1: Nine special characters.
78
Chapter 5
x 1x 2x 3x 4x 5x 6x 7x 8x 9x 10x 11x 12x
0
1
2
3
4
5
6
7
8
9
Γ Ω ˇ Œ ( 2 < F P Z d n x
∆ ↑ ˘ Ø ) 3 = G Q [ e o y
Θ ↓ ¯
Λ ' ˚ ! + 5 ? I S ] g q {
Ξ ¡ ¸ " , 6 @ J T ^ h r 
Π ¿ ß # 7 A K U _ i s }
Σ ı æ $ . 8 B L V ‘ j t ~
Υ œ % / 9 C M W a k u ¨
Φ ` ø & 0 : D N X b l v
Ψ ´ Æ ’ 1 ; E O Y c m w
* 4 > H R \ f p z
Typing text
Table 5.2: Font table for the Computer Modern typewriter style font.
\texttt{\symbol{92}} typesets as \ The argument of the \symbol command is a number matching the slot of the symbol in the layout (encoding) of the font. The layout for the Computer Modern typewriter style font is shown in Table 5.2. Alternatively, instead of \texttt{\symbol{92}}, can use \texttt{\char‘\\} Any character x in the font can be accessed by typing the character itself as ‘\x. This way you don’t have to look up the position of the symbol. You can obtain similar tables for any font in your LATEX implementation by using the fonttbl.tex file in your samples folder. The table format in this file is used in Section 6.6 as an example of the tabular environment. For more about font tables, see the nfssfont.tex file, part of the standard LATEX distribution (see Section 12.3) and see also Section 7.5.7 of The LATEX Companion, 2nd edition [46].
5.4.5
Ellipses The text ellipsis, . . . , is produced using the \dots command. Typing three periods produces ... (notice that the spacing is wrong). \dots is one of several commands that can be used to create ellipses in formulas (see Section 7.4.3).
5.4
79
Symbols not on the keyboard
5.4.6 Ligatures Certain groups of characters, when typeset, are joined together—such compound characters are called ligatures. There are five ligatures that LATEX typesets automatically (if you use the Computer Modern fonts): ff, fi, fl, ffi, and ffl. If you want to prevent LATEX from forming a ligature, separate the characters with the command \textcompwordmark. Compare iff with iff, typed as iff and if\textcompwordmark f Enclosing the second character in braces ({}) is a crude method of preventing the ligature, as used in Formula 4 of the Formula Gallery (see Section 7.9). This method, in some instances, may interfere with LATEX’s hyphenation algorithm.
5.4.7 Accents and symbols in text LATEX provides 15 European accents. Type the command for the accent ( \ and a character), followed by the letter (in braces) on which you want the accent placed (see Table 5.3). For example, to get Gr¨ atzer Gy¨orgy, type Gr\"{a}tzer Gy\"{o}rgy ¨ type \"{O}. and to get O To place an accent on top of an i or a j, you must use the dotless version of i and j. These are obtained by the commands \i and \j: \’{\i} typesets as ´ı and \v{\j} typesets asˇ. Tables 5.4 and 5.5 list some additional text symbols and European characters available in LATEX when typing text. Using localized versions of LATEX, you get more accented and combined characters such as the Catalan geminated ell (see Appendix G). Name
Type
acute breve caron/haˇcek cedilla circumflex dieresis/umlaut double acute grave
\’{o} \u{o} \v{o} \c{c} \^{o} \"{u} \H{o} \‘{o}
dotless i
\i \’{\i}
Typeset
Name
Type
o´ ˘o ˇo ¸c ˆo u ¨ ˝o `o
macron overdot ring tie tilde underdot underbar
\={o} \.{g} \r{u} \t{oo} \~{n} \d{m} \b{o}
¯o g˙ ˚ u o o n ˜ m . o ¯
ı ´ı
dotless j
\j \v{\j}
ˇ
Table 5.3: European accents.
Typeset
80
Chapter 5
Typing text
Note that the \textcircled command (in Table 5.5) takes an argument. It seems a or . α Capitals such as A are to work best with a single lowercase character, like A. not very satisfactory. Section 5.9.6 explains how to create the symbol
5.4.8
Logos and dates \TeX produces TEX, \LaTeX produces LATEX, and \LaTeXe produces LATEX 2ε (the original name of the current version of LATEX). The \AmS command produces the logo AMS. Remember to type \TeX\ or \TeX{} if you need a space after TEX (similarly for the others). A better way to handle this problem is discussed in Section 15.1.1. LATEX also stores some useful numbers: \time is the time of day in minutes since midnight \day is the day of the month \month is the month of the year \year is the current year You can include these numbers in your document by using the \the command: Year: \the\year; month: \the\month; day: \the\day produces a result such as Year: 2006; month: 7; day: 11 Of more interest is the \today command, which produces today’s date in the form: July 11, 2006. It is often used as the argument of the \date command (see Section 11.2.1). Remember the termination rule (Rule 3 in Section 5.3.1). today’s date in the form: \today (you may want produces today’s date in the form: July 11, 2006(you may want To get the desired effect, type \ or {} after the \today command: today’s date in the form: \today\ (you may want
5.4
81
Symbols not on the keyboard Name
Type
Typeset
Type
Typeset
aring aesc ethel eszett inverted question mark inverted exclamation mark slashed L slashed O
\aa \ae \oe \ss ?‘ !‘ \l \o
˚ a æ œ ß ¿ ¡ l ø
\AA \AE \OE \SS
˚ A Æ Œ SS
\L \O
L Ø
Table 5.4: European characters. Name
Type
ampersand asterisk bullet backslash bar (caesura) brace left brace right bullet circled a circumflex copyright dagger double dagger (diesis) dollar double quotation left double quotation right em dash en dash exclamation down greater than less than lowline midpoint octothorp percent pilcrow (paragraph) question down registered trademark section
\& \textasteriskcentered \textbackslash \textbar \{ \} \textbullet \textcircled{a} \textasciicircum \copyright \dag \ddag \$ \textquotedblleft or ‘‘ \textquotedblright or ’’ \textemdash or \textendash or \textexclamdown or !‘ \textgreater \textless \_ \textperiodcentered \# \% \P \textquestiondown or ?‘ \textregistered \S
Table 5.5: Extra text symbols.
Typeset & ∗ \  { } • a
ˆ c
† ‡ $ “ ” — – ¡ > < · # % ¶ ¿ R
§
82
5.4.9
Chapter 5
Typing text
Hyphenation LATEX reads the source file one line at a time until it reaches the end of the current paragraph and then tries to balance the lines (see Section D.3.2). To achieve this goal, LATEX hyphenates long words using a builtin hyphenation algorithm, a database stored in the hyphen.tex file, and a long \hyphenation list in the AMS document classes. If you use a document class not containing such a list, copy the hyphenation list from amsart to your document.
Rule
Optional hyphen If you find that LATEX cannot properly hyphenate a word, put optional hyphens in the word. An optional hyphen is typed as \ , and allows LATEX to hyphenate the word where the optional hyphen is placed—and only at such points—if the need arises. Examples: data\base,an\ti\thet\ic, set\up Note that: Optional hyphens prevent hyphenation at any other point in the word. Placing an optional hyphen in a particular occurrence of a word does not affect the hyphenation of any other occurrences of that word.
Rule
Hyphenation specifications List the words that often need help in a command: \hyphenation{setup associate} All occurrences of the listed words following this command in your document are hyphenated as specified. Note that in the \hyphenation command the hyphens are designated by  and not by \, and that the words are separated by spaces not by commas. You must use optional hyphens for words with accented characters, as in Gr\"{a}t\zer Such words cannot be included in a \hyphenation list (unless you use the T1 font encoding—see Appendix G).
5.4
83
Symbols not on the keyboard
Rule
Preventing hyphenation To prevent hyphenation of a word, put it in the argument of a \text command or place it unhyphenated in a \hyphenation command. For example, type \text{database} if you do not want this instance of database hyphenated, or type \hyphenation{database} if you do not want LATEX to hyphenate any occurrence of the word after this command in your document. Of course, typing data\base overrides the general prohibition for this one instance. You can have any number of \hyphenation commands in your document.
Tip
LATEX does not break a hyphenated word except at the hyphen, nor does it break a word followed by or preceding an em dash or en dash (see Section 5.4.2). LATEX often needs help with such words. Sometimes a hyphen in a phrase should not be broken. For instance, the phrase mcomplete lattice should not be broken after m; so type it as \text{$\mathfrak{m}$com}\plete lattice (see Section 8.3.2 for \mathfrak). Use the \nobreakdash command (placed before the hyphen) \nobreakdash
\nobreakdash
\nobreakdash
to prevent such breaks. For example, pages~24\nobreakdash47 Since LATEX does not hyphenate a hyphenated word except at the hyphen, \nobreakdashprevents the hyphenation of the whole word as though it were enclosed in a \text command. The form \nobreakdash\hspace{0pt} allows the normal hyphenation of the word that follows the hyphen. For example,
84
Chapter 5
Typing text
$\mathfrak{m}$\nobreakdash\hspace{0pt}complete lattice allows the word complete to be hyphenated. This coding of the phrase mcomplete lattice is a natural candidate for a userdefined command (see Section 15.1.1).
Tip
If you want to know how LATEX would hyphenate a list of words, place it in the argument of a \showhyphens command. For instance, \showhyphens{summation reducible latticoid} The result, summation reducible latticoid is shown in the log file.
Tip
Some text editors wrap lines in a source file by breaking them at a hyphen, introducing errors in your typeset document. For instance, It follows from Theorem~\ref{T:M} that completesimple lattices are very large. is typeset by LATEX as follows: It follows from Theorem 2 that complete simple lattices are very large. As you can see, there is a space between the hyphen and the word simple. The text editor inserted an endofline character after the hyphen (by the second space rule, see Section 5.2.1). This endofline character was interpreted by LATEX as a space. To correct the error, make sure that there is no such line break, or comment out (see Section 5.5.1) the endofline character: It follows from Theorem~\ref{T:M} that complete% simple lattices are very large. Better yet, rearrange the two lines: It follows from Theorem~\ref{T:M} that completesimple lattices are very large.
5.5
Comments and footnotes
85
Of course, LATEX does not know everything about the complicated hyphenation rules of the English language. Consult The Chicago Manual of Style, 15th edition [11] and Lyn Dupr´e’s BUGS in Writing: A Guide to Debugging Your Prose, 2nd edition [13] for additional guidance.
5.5
Comments and footnotes Various parts of your source file do not get typeset like most of the rest. The two primary examples are comments that do not get typeset at all and footnotes that get typeset at the bottom of the page.
5.5.1 Comments The % symbol tells LATEX to ignore the rest of the line. A common use might be a comment to yourself in the source file: therefore, a reference to Theorem~1 % check this! The % symbol has many uses. For instance, a document class command (see Section 11.5), \documentclass[twocolumn,twoside,legalpaper]{amsart} may be typed with explanations, as \documentclass[% twocolumn,% option for twocolumn pages twoside,% format for twosided printing legalpaper% print on legalsize paper ]{amsart} so you can easily comment out some at a later time, as in \documentclass[% %twocolumn,% option for twocolumn pages %twoside,% format for twosided printing legalpaper% print on legalsize paper ]{amsart} Notice that the first line is terminated with a % to comment out the endofline character.
Tip
Some command arguments do not allow any spaces. If you want to break a line within an argument list, you can terminate the line with a %, as shown in the previous example. See also the example at the end of Section 5.4.9.
86
Chapter 5
Typing text
It is often useful to start a document with a comment line giving the file name and identifying the earliest version of LATEX that must be used to typeset it. % This is article.tex \NeedsTeXFormat{LaTeX2e}[1994/12/01] The second line specifies the December 1, 1994 (or later) release of LATEX. You may need to use such a declaration if your document uses a feature that was not available in earlier releases. Since LATEX changes very little these days, this command is of limited use. (See, however, the discussion on page 311.) Other uses of % include marking parts of the article for your own reference. For instance, you may include comments to explain command definitions (as in Section 15.3). If something goes wrong inside a multiline math display (see Chapter 9), LATEX does not tell you precisely where the error is. You can try commenting out all but one of the lines, until each line works separately. Note that % does not comment out lines in a B IBTEX database document (see Section 16.2.4).
Tip
The 25% rule If you want a % sign in text, make sure you type it as \%. Otherwise, % comments out the rest of the line. LATEX does not produce a warning. Using % to comment out large blocks of text can be tedious even with block comment. The verbatim package includes the comment environment: \begin{comment} ...the commented out text... \end{comment}
Rule
comment environments 1. \end{comment} must be on a line by itself. 2. There can be no comment within a comment. In other words, \begin{comment} commented out text... \begin{comment} some more commented out text... \end{comment}
5.5
Comments and footnotes
87
and some more commented out text... \end{comment} is not allowed. LATEX may give one of several error messages, depending on the circumstances. For instance, ! LaTeX Error: \begin{document} ended by \end{comment}. l.175 \end{comment} The comment environment can be very useful in locating errors. Suppose you have unbalanced braces in your source file (see Section 5.3.2). Working with a copy of your source file, comment out the first half at a safe point (not within an environment!) and typeset. If you still get the same error message, the error is in the second half. If there is no error message, the error is in the first half. Comment out the half that has no error. Now comment out half of the remaining text and typeset again. Check to see whether the error appears in the first half of the remaining text or the second. Continue applying this method until you narrow down the error to a paragraph that you can inspect visually. Since the comment environment requires the verbatim package, you must include the line \usepackage{verbatim} in the preamble of the source file (see Sections 12.3.1 and E.1).
5.5.2 Footnotes The text of a footnote is typed as the argument of a \footnote command. To illustrate the use of footnotes, I have placed one here.1 This footnote is typed as \footnote{Footnotes are easy to place.} If you want to use symbols to designate the footnotes, instead of numbers, type the command \renewcommand{\thefootnote} {\ensuremath{\fnsymbol{footnote}}} before the first footnote; this provides up to nine symbols. In Section 15.1.1, we discuss the \ensuremath command. Section 3.2 of The LATEX Companion, 2nd edition [46] describes how to further customize footnotes. In addition, there are titlepage footnotes, such as the \thanks and \date commands in the top matter. See page 42 for a typeset example of \date. See also Section 11.2 and the typeset title page footnotes on page 286. 1 Footnotes
are easy to place.
88
Chapter 5 You can add a footnote marked by the title
∗
Typing text
to the title of an article. For instance, type
\title[Complete congruence lattices]% {Complete congruence lattices$^*$} and add the lines {\renewcommand{\thefootnote}{\fnsymbol{footnote}} \setcounter{footnote}{1} \footnotetext{Lecture delivered at the \AMS annual meeting in Brandon.} \setcounter{footnote}{0} } The footnote will appear as the first footnote on page 1 marked by ∗ . All the other footnotes are unmarked.
5.6
Changing font characteristics Although a document class and its options determine how LATEX typesets characters, there are occasions when you want control over the shape or size of the font used.
5.6.1
Basic font characteristics You do not have to be a typesetting expert to recognize the following basic font attributes: Shape Normal text is typeset: upright (or roman) as this text slanted as this text italic as this text small caps as this text Monospaced and proportional Typewriters used monospaced fonts, that is, fonts all of whose characters are of the same width. Most text editors display text using a monospaced font. LATEX calls monospaced fonts typewriter style. In this book, such a font is used to represent user input and LATEX’s response, such as “typewriter style text”. Whereas, normal text is typeset in a proportional font, such as “proportional text with ii and mm”, in which i is narrow and m is wide: mmmmmm monospaced iiiiii mmmmmm iiiiii
proportional
5.6
Changing font characteristics
89
Serifs A serif is a small horizontal (sometimes vertical) stroke used to finish off a vertical stroke of a letter, as on the top and bottom of the letter M. LATEX’s standard serif font is Computer Modern roman, such as “serif text”. Fonts without serifs are called sans serif, such as “sans serif text”. Sans serif fonts are often used for titles or for special emphasis. Series: weight and width The series is the combination of weight and width. A font’s weight is the thickness of the strokes and the width is how wide the characters are. Light, medium (or normal), and bold often describe weight. Narrow (or condensed), medium (or normal), and extended often describe width. The Computer Modern family includes bold fonts. Traditionally, when the user asks for bold CM fonts, LATEX actually provides bold extended (a somewhat wider version). Size Most LATEX articles are typeset with 10 point text unless otherwise instructed. Larger sizes are used for titles, section titles, and so on. Abstracts and footnotes are often set in 8point type. Font family The collections of all sizes of a font is called a font family.
5.6.2 Document font families In a document class, the style designer designates three document font families: 1. Roman (upright and serifed) document font family 2. Sans serif document font family 3. Typewriter style document font family and picks one of these (for articles, as a rule, the roman document font family) as the document font family or normal family. In all the examples in this book, the document font family is the roman document font family except for presentations which use sans serif (see Section 4.4 and Chapter 14). When you use Computer Modern fonts in LATEX, which is the default, the three document font families are Computer Modern roman, Computer Modern sans serif, and Computer Modern typewriter. The document font family is Computer Modern roman. In this book, the roman document font family is Times, the sans serif document font family is Helvetica, and the typewriter style document font family is Computer Modern typewriter. The document font family is the roman document font family Times. The document font family (normal family) is the default font. You can always switch back to it with \textnormal{...} or {\normalfont ...}
90
Chapter 5
Typing text
Table 5.6 shows these two commands and three additional pairs of commands to help you switch among the three basic document font families. It also shows the command pairs for the basic font shapes. Command with Argument
Command Declaration
Switches to the font family
\textnormal{...} \emph{...} \textrm{...} \textsf{...} \texttt{...} \textup{...} \textit{...} \textsl{...} \textsc{...} \textbf{...} \textmd{...}
{\normalfont ...} {\em ...} {\rmfamily ...} {\sffamily ...} {\ttfamily ...} {\upshape ...} {\itshape ...} {\slshape ...} {\scshape ...} {\bfseries ...} {\mdseries ...}
document emphasis roman sans serif typewriter style upright shape italic shape slanted shape small capitals bold normal weight and width
Table 5.6: Font family switching commands.
Command pairs The fontchanging commands of Table 5.6 come in two forms: A command with an argument, such as \textrm{...}, changes its argument. These are short commands, i.e., they cannot contain a blank line or a \par command. A command declaration, such as \rmfamily, carries out the font change following the command and within its scope (see Section 5.3.2). You should always use commands with arguments for small changes within a paragraph. They have two advantages: You are less likely to forget to change back to the normal font. You do not have to worry about italic corrections (see Section 5.6.4). Note that MakeIndex requires you to use commands with arguments to change the font in which page numbers are typeset (see Section 17.1). For font changes involving more than one paragraph, use command declarations. These commands are preferred if you want to create userdefined commands and environments (see Chapter 15).
5.6.3
Shape commands There are five pairs of commands to change the font shape:
5.6
Changing font characteristics
91
\textup{...} or {\upshape ...} switch to the upright shape. \textit{...} or {\itshape ...} switch to the italic shape. \textsl{...} or {\slshape ...} switch to the slanted shape. \textsc{...} or {\scshape ...} switch to small capitals. \emph{...} or {\em ...} switch to emphasis. The document class specifies how emphasis is typeset. As a rule, it is italic or slanted unless the surrounding text is italic or slanted, in which case it is upright. For instance, \emph{Rubin space} in the statement of a theorem is typeset as the space satisfies all three conditions, a socalled Rubin space that ... The emphasis changed the style of Rubin space from italic to upright.
Tip
Rule
Be careful not to interchange the command pairs. For instance, if by mistake you type {\textit serif}, the result is serif. Only the s is italicized since \textit takes s as its argument.
Abbreviations and acronyms For abbreviations and acronyms use small caps, except for twoletter geographical acronyms. So Submitted to tug should be typed as Submitted to \textsc{tug} Note that only the lowercase characters in the argument of the \textsc command are printed as small caps.
5.6.4 Italic corrections The phrase when using a serif font may be typed as follows:
92
Chapter 5
Typing text
when using a {\itshape serif\/} font The \/ command before the closing brace is called an italic correction. Notice that {\itshape M}M typesets as MM, where the M is leaning into the M. Type {\itshape M\/}M to get the correct spacing M M. Compare the typeset phrase from the previous example with and without an italic correction: when using a serif font when using a serif font The latter is not as pleasing to the eye.
Rule 1
Italic correction If the emphasized text is followed by a period or comma, you should not type the italic correction. For example, Do not forget. My party is on Monday. should be typed as {\itshape Do not forget.} My party is on Monday.
Rule 2
Italic correction The shape commands with arguments do not require italic correction. The corrections are provided automatically where needed. Thus you can type the phrase when using a serif font the easy way: when using a \textit{serif} font Whenever possible, let LATEX take care of the italic correction. However, if LATEX is adding an italic correction where you feel it is not needed, you can override the correction with the \nocorr command. LATEX does not add an italic correction before a period or a comma. These two punctuation marks are stored in the \nocorrlist command. By redefining this command, you can modify LATEX’s behavior.
Rule 3
Italic correction The italic correction is required with the commands \itshape, \slshape, \em.
5.6
93
Changing font characteristics Command
Sample text
\Tiny \tiny \SMALL or \scriptsize \Small or \footnotesize \small
sample text
\normalsize
sample text
sample text
sample text
sample text
sample text
sample text
\large \Large
sample text
sample text
\LARGE
sample text
\huge
sample text
\Huge
Table 5.7: Font size commands.
5.6.5 Series These attributes play a very limited role with the Computer Modern fonts. There is only one important pair of commands, \textbf{...}
{\bfseries ...}
to change the font to bold (actually, bold extended). The commands \textmd{...}
{\mdseries ...}
which set both the weight and width to medium (normal) are seldom needed.
5.6.6 Size changes Standard LATEX documents are typeset in 10 point type. The 11 point and 12 point type are often used for greater readability and some journals require 12 point—if this is the case, use the 12pt document class option (see Sections 11.5, 12.1.2, and 18.1.3). The sizes of titles, subscripts, and superscripts are automatically set by the document class, in accordance with the font size option. If you must change the font size for some text—it is seldom necessary to do so in an article—the following command declarations are provided (see Table 5.7): \Tiny
\tiny
\SMALL
\Small
\small
\normalsize \large
\Large
\LARGE
\huge
\Huge
94
Chapter 5
Typing text
The command \SMALL is also called \scriptsize and the command \Small is also called \footnotesize. The font size commands are listed in order of increasing—to be more precise, nondecreasing—size. Two commands allow the user to increase or decrease font size: \larger moves up one size, \smaller moves down one. Both commands take an optional argument. For example, \larger[2] moves up 2 sizes.
5.6.7
Orthogonality You are now familiar with the commands that change the font family, shape, series, and size. Each of these commands affects one and only one font attribute. For example, if you change the series, then the font family, shape, and size do not change. These commands act independently. In LATEX terminology, the commands are orthogonal. From the user’s point of view this behavior has an important consequence: The order in which these commands are given does not matter. Thus \Large \itshape \bfseries has the same effect as \bfseries \itshape \Large Note that LATEX 2.09’s twoletter commands (see Section 5.6.8) are not orthogonal. Orthogonality also means that you can combine these font attributes in any way you like. For instance, the commands \sffamily \slshape \bfseries \Large instruct LATEX to change the font family to sans serif, the shape to slanted, the series to bold, and the size to \Large. If the corresponding font is not available, LATEX uses a font that is available, and issues a warning. The font substitution algorithm (see Section 7.9.3 of The LATEX Companion, 2nd edition [46] for details) may not provide the font you really want, so it is your responsibility to make sure that the necessary fonts are available. We discuss this topic further in Section 18.5.
5.6.8
Obsolete twoletter commands Users of LATEX 2.09 and AMSLATEX version 1.1 are accustomed to using the twoletter commands \bf, \it, \rm, \sc, \sf, \sl, and \tt. These commands are not part of LATEX. They are, however, still defined in most document classes. The twoletter commands 1. switch to the document font family, 2. change to the requested shape. There are a number of reasons not to use them. The twoletter commands
5.7
Lines, paragraphs, and pages
95
are not part of LATEX, require manual italic corrections, are not orthogonal (see Section 5.6.7). \slshape \bfseries is the same as \bfseries \slshape (slanted bold), but \sl\bf is not the same as \bf\sl. Indeed, {\sl\bf sample} gives sample and {\bf\sl sample} produces sample.
5.6.9 Lowlevel commands The fontcharacteristic changing commands we discussed in this section are the highlevel font commands. Each of these commands is implemented by LATEX and the document class using lowlevel font commands. The lowlevel commands have been developed for document class and package writers. See Section 7.9 of The LATEX Companion, 2nd edition [46]. There is one use of lowlevel commands you should keep in mind. When you choose a font size for your document or for some part thereof, you also determine the \baselineskip, the distance from the baseline of one line to the baseline of the next. Typically, a 10point font size uses a 12 point \baselineskip. Occasionally, you may want to change the font size along with the \baselineskip. A command for accomplishing this is \fontsize{9pt}{11pt}\selectfont which changes the font size to 9 point and the \baselineskip to 11 point. To make this change for a single paragraph, you can type {%special paragraph \fontsize{9pt}{11pt}\selectfont text }%end special paragraph Observe the blank line that follows text and marks the end of the paragraph; \par would accomplish the same thing.
5.7
Lines, paragraphs, and pages When typesetting a document, LATEX breaks the text into lines, paragraphs, and pages. Sometimes you may not like how LATEX has chosen to lay out your text. There are ways to influence how LATEX does its work and these are discussed in this section.
96
Chapter 5
5.7.1
Typing text
Lines LATEX typesets a document one paragraph at a time. It tries to split the paragraph into lines of equal width. If it fails to do so successfully and a line is too wide, you get an overfull \hbox message. Here is a typical example: Overfull \hbox (15.38948pt too wide) in paragraph at lines 1116 []\OT1/cmr/m/n/10 In several sections of the course on matrix theory, the strange term ‘‘hamiltonianThe log file records these error messages. To place a visual warning in the typeset version of your document as well, use the draft document class option \documentclass[draft]{amsart} Lines that are too wide are be marked with a slug (a black box) in the margin. A slug is a vertical bar of width \overfullrule. Do not worry about such messages while writing the document. If you are preparing the final version and receive a message for an overfull \hbox, the first line of defense is to see whether optional hyphens would help (see Section 5.4.9). Read the warning message carefully to see which words LATEX cannot hyphenate properly. If adding optional hyphens does not help, a simple rephrasing of the problem sentence often does the trick. Recall that there are 72.27 points in an inch (see Section 2.3). So if the error message indicates a 1.55812 pt overflow, for instance, you can safely ignore it.
Tip
If you do not want the 1.55812pt overflow reported whenever the document is typeset, you can enclose the offending paragraph (including the blank line indicating the end of the paragraph) between the lines {\setlength{\hfuzz}{2pt} and }% end of \hfuzz=2pt Choose an argument that is slightly more than the reported error (maybe 2pt). This does not affect the typeset output, but the warning message and the slug, if you are using the draft option, are suppressed. Alternatively, enclose the offending paragraph including the blank line indicating the end of the paragraph in a setlength environment:
5.7
Lines, paragraphs, and pages
97
\begin{setlength}{\hfuzz}{2pt} \end{setlength} Breaking lines There are two forms of the line breaking command: The \\ and \newline commands break the line at the point of insertion but do not stretch it. The \linebreak command breaks the line at the point of insertion and stretches the line to make it of the normal width. The text following any of these commands starts at the beginning of the next line, without indentation. The \\ command is often used, but \linebreak is rarely seen. (See Section 18.6 for an application of the \linebreak command.) I illustrate the effect of these commands: There are two forms of the line breaking command: There are two forms\\ of the line breaking command: There are two forms \newline of the line breaking command: There are two forms \linebreak of the line breaking command: typeset as There are two forms of the line breaking command: There are two forms of the line breaking command: There are two forms of the line breaking command: There are two of the line breaking command:
forms
If you force a line break in the middle of a paragraph with the \linebreak command and LATEX thinks that there is too little text left on the line to stretch it to full width, you get a message such as Underfull \hbox (badness 4328) in paragraph at lines 812
98
Chapter 5
Typing text
The \\ command has two important variants: \\[length ], where length is the interline space you wish to specify after the line break, for instance, 12pt, .5in, or 1.2cm. Note how the units are abbreviated. \\*, which prohibits a page break following the line break. The \\*[length ] form combines the two variants. We illustrate the \\[length ] command: It is also semimodular.\\[15pt]
In particular,
which is typeset as It is also semimodular. In particular, Since \\ can be modified by * or by [ ], LATEX may get confused if the line after a \\ command starts with a * or [. In such cases, type * as {*} or [ as {[}. For instance, to get There are two sources of problems: [a] The next line starts with [. type There are two sources of problems:\\ {[}a] The next line starts with \texttt{[}. If you fail to type {[}, you get the error message ! Missing number, treated as zero. a l.16 [a] The next line starts with \texttt{[}.
Rule
\\ Without optional arguments, the \\ command and the \newline command are the same in text, but not within environments or command arguments.
5.7
Lines, paragraphs, and pages
99
You can qualify the \linebreak command with an optional argument: 0 to 4. The higher the argument, the more it forces the occurrence of a line break. The \linebreak[4] command is the same as \linebreak, while \linebreak[0] allows the line break but does not force it. The \nolinebreak command plays the opposite role. \nolinebreak[0] = \linebreak[0], and \nolinebreak[4] = \nolinebreak. \nolinebreak is seldom used since the tie (~) and the \text command (see Section 5.4.3) accomplish the same goal most of the time. Double spacing It is convenient to proofread documents double spaced. Moreover, some journals require submissions to be double spaced. To typeset a document double spaced, include the command \renewcommand{\baselinestretch}{1.5} in its preamble. Alternatively, get George D. Greenwade’s setspace (see Section E.1 on how to get it). Load this package with a \usepackage{setspace} command in the preamble of the document and specify \doublespacing in the preamble. This changes not just the line spacing but a number of other parameters to make your article look good. See also Section 3.1.13 of The LATEX Companion, 2nd edition [46].
5.7.2 Paragraphs Paragraphs are separated by blank lines or by the \par command. Error messages always show paragraph breaks as \par. The \par form is also very useful in userdefined commands and environments (see Sections 15.1 and 15.2). In some document classes, the first line of a paragraph is automatically indented. Indentation can be prevented with the \noindent command and can be forced with the \indent command. Sometimes—for instance, in a schedule, glossary, or index—you may want a hanging indent, where the first line of a paragraph is not indented, and all the others are indented by a specified amount. Hanging indents are created by specifying the amount of indentation specified by \hangindent and set with the \setlength command: \setlength{\hangindent}{30pt} \noindent
100
Chapter 5
Typing text
\textbf{sentence} a group of words terminated by a period, exclamation point, or question mark. \setlength{\hangindent}{30pt} \noindent \textbf{paragraph} a group of sentences terminated by a blank line or by the new paragraph command. produces sentence a group of words terminated by a period, exclamation point, or question mark. paragraph a group of sentences terminated by a blank line or by the new paragraph command. Notice that the \setlength command must be repeated for each paragraph. Sometimes you may want to change the value of \hangafter, the length command that specifies the number of lines not to be indented. The default value is 1. To change it to 2, use the command \setlength{\hangafter}{2} For more about the \setlength command, see Section 15.5.2. The LATEX Companion, 2nd edition [46] discusses in Section 3.1.4 the style parameters of a paragraph. The preferred way to shape a paragraph or series of paragraphs is with a custom list environment (see Section 15.6).
5.7.3
Pages There are two page breaking commands: \newpage, which breaks the page at the point of insertion but does not stretch the content \pagebreak, which breaks the page at the point of insertion and stretches the page’s content to normal length Text following either command starts at the beginning of the next page, indented. As you can see, the page breaking commands are analogous to the line breaking commands discussed in Section 5.7.1. This analogy continues with the optional argument, 0 to 4: \pagebreak[0] to \pagebreak[4] \nopagebreak[0] to \nopagebreak[4]
5.7
Lines, paragraphs, and pages
101
There are also special commands for allowing or forbidding page breaks in multiline math displays (see Section 9.9). When preparing the final version of a document (see Section 18.5), you may have to extend or shrink a page by a line or two to prevent it from breaking at an unsuitable line. You can do so with the \enlargethispage command. For instance, \enlargethispage{\baselineskip} adds one line to the page length. On the other hand, \enlargethispage{\baselineskip} makes the page one line shorter. \enlargethispage{10000pt} makes the page very long. The *ed version, \enlargethispage*, squeezes the page as much as possible. There are two more variants of the \newpage command. The \clearpage command does a \newpage and typesets all the figures and tables waiting to be processed (see Section 10.4.3). The variant \cleardoublepage is used with the twoside document class option (see Sections 11.5 and 12.1.2). It does a \clearpage and in addition makes the next printed page a righthand, that is, oddnumbered, page, by inserting a blank page if necessary. If for your document class this does not work, use the package cleardoublepage.sty in the samples folder. Section 18.6 discusses the use of some of these commands in the final preparation of books.
5.7.4 Multicolumn printing Many document classes provide the twocolumn option for twocolumn typesetting (see Sections 11.5 and 12.1.2). In addition, there is a \twocolumn command which starts a new page by issuing a \clearpage and then typesets in two columns. An optional argument provides a twocolumn wide title. Use the \onecolumn command to switch back to a onecolumn format. Frank Mittelbach’s multicol package (see Section 12.3.1) provides the much more sophisticated multicols environment, which can start in the middle of a page, can handle more than two columns, and can be customized in a number of ways (see Section 3.5.3 of The LATEX Companion, 2nd edition [46]).
102
5.8
Chapter 5
Typing text
Spaces The judicious use of horizontal and vertical space is an important part of the formatting of a document. Fortunately, most of the spacing decisions are made by the document class, but LATEX has a large number of commands that allow the user to insert horizontal and vertical spacing. Remember that LATEX ignores excess spaces, tabs, and endofline characters. If you need to add horizontal or vertical space, then you must choose from the commands in this section. Use them sparingly.
5.8.1
Horizontal spaces In this section, we discuss fixed length horizontal space commands. Variable length horizontal space is discussed in Section 5.8.4. When typing text, there are three commands that are often used to create horizontal space, shown between the bars in the display below: \ \quad \qquad The \quad command creates a 1 em space and \qquad creates a 2 em space (see Section 5.8.3). The interword space created by \ can both stretch and shrink. There are other commands that create smaller amounts of space. All the math spacing commands of Section 8.1.3—with the exception of \mspace—can be used in ordinary text (see Sections B.9 and C.6), but the \hspace and \phantom commands are more appropriate. The \hspace command takes a length as a parameter. The length may be negative. For example, \textbar\hspace{12pt}\textbar \textbar\hspace{.5in}\textbar \textbar\hspace{1.5cm}\textbar
  



or \hspace{40pt}. The command \hspace is often used with a negative argument when placing illustrations. The \phantom{argument } command produces a space the width and height of the space that would be occupied by its typeset argument \textbar need space\textbar \textbar\phantom{need space}\textbar and alpha \phantom{beta} gamma \phantom{delta}\\ \phantom{alpha} beta \phantom{gamma} delta
need space  
5.8
103
Spaces produces alpha
gamma beta
delta
The \phantom command is very useful for fine tuning aligned math formulas (see Sections 8.1 and 9.5.3). The variant \hphantom{argument } creates a space with the horizontal dimension that would be occupied by its typeset argument and with zero height. For instance, the last two lines of the dedication of this book were typed as follows: \textbf{Emma} (2), \phantom{and \textbf{Kate} (0)}\\[8pt] and \textbf{Kate} (0) See Section C.6 for a table of all horizontal textspacing commands. Horizontal space variant When LATEX typesets a line, it removes all spaces from the beginning of the line, including the space created by \hspace, \quad, and other spacing commands. Using the *ed variant of \hspace, \hspace*, prevents LATEX from removing the space you have specified. For example, And text\\ \hspace{20pt}And text\\ \hspace*{20pt}And text is typeset as And text And text And text Use the \hspace* command for creating customized indentation. To indent a paragraph by 24 points, give the command \noindent\hspace*{24pt}And text which typesets as
104
Chapter 5
Typing text
And text To break a line and indent the next line by 24 points, give the command And text\\ \hspace*{24pt}And text which produces And text And text
5.8.2
Vertical spaces You can add some interline space with the command \\[length ], as discussed in Section 5.7.1. You can also do it with the \vspace command, which works just like the \hspace command (see Section 5.8.1), except that it creates vertical space. Here are some examples: \vspace{12pt}
\vspace{.5in}
\vspace{1.5cm}.
Standard amounts of vertical space are provided by the three commands \smallskip
\medskip
\bigskip
The space these commands create depends on the document class and the font size. With the document class and font I am using for this book, they represent a vertical space of 3 points, 6 points, and 12 points, respectively. 12 points is the baseline skip (see Section 5.6.9) in standard LATEX documents with the default 10pt option.
Rule
Vertical space commands All vertical space commands add the vertical space after the typeset line in which the command appears. To obtain end of text. New paragraph after vertical space
5.8
105
Spaces type end of text. \vspace{12pt} New paragraph after vertical space
The following example illustrates the unexpected way the vertical space is placed if the command that creates it does not start a new paragraph: end of text. \vspace{12pt} The following example illustrates the unexpected way the vertical space is placed if the command that creates it does not start a new paragraph: It typesets as end of text. The following example illustrates the unexpected way the vertical space is placed if the command that creates it does not start a new paragraph:
Vertical space variants LATEX removes vertical space from the beginning and end of each page, including space produced by \vspace. The space created by the variant \vspace* is not removed by LATEX under any circumstances. Use this command, for instance, to start the typeset text (say, of a letter) not at the top of the page. The \phantom command has also a vertical variant: \vphantom. The command \vphantom{argument } creates a vertical space with the vertical dimension that would be occupied by its typeset argument, argument .
5.8.3 Relative spaces The length of a space is usually given in absolute units: 12pt (points), .5cm (centimeters), 1.5in (inches). Sometimes, relative units, em and ex, are more appropriate, units that are relative to the size of the letters in the current font. The unit 1 em is approximately the width of an M in the current font, 1 ex is approximately the height of an x in the current font. These units are used in commands such as \hspace{12em} and \vspace{12ex} The \quad and \qquad commands (Section 5.8.1) produce 1 em and 2 em spaces.
106
5.8.4
Chapter 5
Typing text
Expanding spaces Horizontal spaces The \hfill, \dotfill, and \hrulefill commands fill all available space in the line with spaces, dots, or a horizontal line, respectively. If there are two of these commands on the same line, the space is divided equally between them. These commands can be used to center text, to fill lines with dots in a table of contents, and so on. To obtain 2. Boxes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34 ABC and DEF ABC and DEF type 2. Boxes\dotfill 34\\ ABC\hfill and\hfill DEF\\ ABC\hrulefill and\hrulefill DEF In a centered environment—such as a \title (see Section 11.2.1) or a center environment (see Section 6.3)—you can use \hfill to set a line flush right: This is the title
First Draft
Author To achieve this effect, type \begin{center} This is the title\\ \hfill First Draft\\ Author \end{center} Vertical spaces The vertical analogue of \hfill is \vfill. This command fills the page with vertical space so that the text before the command and the text after the command stretch to the upper and lower margin. You can play the same games with it as with \hfill in Section 5.8.4. The command \vfill stands for \vspace{\fill}, so it is ignored at the beginning of a page. Use \vspace*{\fill} if you need it at the beginning of a page.
5.9
5.9
107
Boxes
Boxes Sometimes it can be useful to typeset text in an imaginary box, and treat that box as a single large character. A singleline box can be created with the \text or \makebox commands and a multiline box of a prescribed width can be created with the \parbox command or minipage environment.
5.9.1 Line boxes The \text command provides a line box that typesets its argument without line breaks. As a result, you may find the argument extending into the margin. The resulting box is handled by LATEX as if it were a single large character. For instance, \text{database} causes LATEX to treat the eight characters of the word database as if they were one. This technique has a number of uses. It prevents LATEX from braking the argument (see Section 5.4.9). It also allows you to use the phrase in the argument in a formula (see Section 7.4.6). The argument of \text is typeset in a size appropriate for its use, for example, as a subscript or superscript. See Section 7.4.6 for an example. Line boxes—a refinement The \mbox command is the short form of the \makebox command. Both \mbox and \text prevent breaking the argument, but \mbox does not change size in subscripts and superscripts. The full form of the \makebox command is \makebox[width ][alignment ]{text } where the arguments are width, the (optional) width of the box. If [width ] is omitted, the box is as wide as necessary to enclose its contents. alignment, (optionally) one of c (the default), l, r, or s. The text is centered by default, l sets the argument flush left, r right, and s stretches the text the full length of the box if there is blank space in the argument. text, the text in the box. A width argument can be specified in inches (in), centimeters (cm), points (pt), or relative units such as em or ex (see Sections 5.8.3 and 15.5.2).
108
Chapter 5
Typing text
The following examples, \makebox{Short title.}End\\ \makebox[2in][l]{Short title.}End\\ \makebox[2in]{Short title.}End\\ \makebox[2in][r]{Short title.}End\\ \makebox[2in][s]{Short title.}End typeset as Short title.End Short title. Short title. Short
End End Short title.End title.End
The optional width argument, width, can use four length commands: \height
\depth
\totalheight
and \width
These are the dimensions of the box that would be produced without the optional width argument. Here is a simple example. The command \makebox{hello} makes a box of width \width. To typeset hello in a box three times the width, that is, in a box of width 3\width, use the command \makebox[3\width]{hello} So start\makebox[3\width]{hello}end typesets as start
hello
end
The formal definition of these four length commands is the following: \height is the height of the box above the baseline \depth is the depth of the box below the baseline \totalheight is the sum of \height and \depth \width is the width of the box
5.9
109
Boxes
There is an interesting variant of the \makebox command. The \rlap command makes a box and pretends that it is of width zero. For instance, \newcommand{\circwithdot} {\mathbin{\rlap{$\mspace{2mu}\cdot$}\hbox{$\circ$}}} defines the command \circwithdot, so you can type $f\circwithdot\varphi$ which prints as f ◦· ϕ. There is also an \llap command.
5.9.2 Frame boxes Boxed text is very emphatic. For example, Do not touch! is typed as \fbox{Do not touch!} This is a frame box, hence the command \fbox or \framebox. Boxed text cannot be broken, so if you want a frame around more than one line of text, you should put the text as the argument of a \parbox command or within a minipage environment (see Section 5.9.3), and then put that into the argument of an \fbox command. For instance, \fbox{\parbox{3in}{Boxed text cannot be broken, so if you want to frame more than one line of text, place it in the argument of a \bsl\texttt{parbox} command or within a \texttt{minipage} environment.}} produces Boxed text cannot be broken, so if you want to frame more than one line of text, place it in the argument of a \parbox command or within a minipage environment. The \bsl command is defined in Section 15.1.1. See Section 8.6.2 for boxed formulas. The \framebox command works exactly like \makebox, except that it draws a frame around the box. \framebox[2in][l]{Short title}
110
Chapter 5
Typing text
produces Short title You can use this command to typeset the number 1 in a square box, as required by the title of Michael Doob’s TEX Starting from 1 [12]: \framebox{\makebox[\totalheight]{1}} which typesets as 1 Note that \framebox[\totalheight]{1} typesets as 1 which is not a square box. Indeed, \totalheight is the height of 1, which becomes the width of the box. The total height of the box, however, is the height of the character 1 to which you have to add twice the \fboxsep, the separation between the contents of the box and the frame, defined as 3 points, and twice the \fboxrule, the width of the line, or rule, defined as 0.4 points. These lengths are in general also added to the width of the box, but not in this case, because we forced the width to equal the height of the character. You can use the \fbox command to frame the name of an author: \author{\fbox{author’s name}}
5.9.3
Paragraph boxes A paragraph box works like a paragraph. The text it contains is wrapped around into lines. The width of these lines is set by the user. The \parbox command typesets the contents of its second argument as a paragraph with a line width supplied as the first argument. The resulting box is handled by LATEX as a single large character. For example, to create a 3inch wide column, Fred Wehrung’s new result shows the limitation of E. T. Schmidt’s construction, especially for large lattices.
5.9
111
Boxes type \parbox{3in}{Fred Wehrung’s new result shows the limitation of E.\,T. Schmidt’s construction, especially for large lattices.}
Paragraph boxes are especially useful when working within a tabular environment. See the subsection on refinements in Section 6.6 for examples of multiline entries. The width of the paragraph box can be specified in inches (in), centimeters (cm), points (pt), or the relative measurements em and ex (see Section 5.8.3), among others (see Section 15.5.2 for a complete listing of measurement units).
Tip
The \parbox command requires two arguments. Dropping the first argument results in an error message such as ! Missing number, treated as zero. T l.175 Dropping the second argument does not yield an error message but the result is probably not what you intended. The next character is taken as the contents of the \parbox.
Paragraph box refinements The “character” created by a \parbox command is placed on the line so that its vertical center is aligned with the center of the line. An optional first argument b or t forces the paragraph box to align along its bottom or top. For an example, see Section 6.6. The full syntax of \parbox is \parbox[alignment ][height ][inneralignment ]{width }{text } Just as for the \makebox command (see Section 5.9.1), the \height
\depth \totalheight and
\width
commands may be used in the height argument instead of a numeric argument. The inneralignment argument is the vertical equivalent of the alignment argument for \makebox, determining the position of text within the box and it may be any one of t, b, c, or s, denoting top, bottom, centered, or stretched alignment, respectively. When the inneralignment argument is not specified, it defaults to alignment.
112
Chapter 5
Typing text
Paragraph box as an environment The minipage environment is very similar to the \parbox command. It typesets the text in its body using a line width supplied as an argument. It has an optional argument for bottom or top alignment, and the other \parbox refinements also apply. The difference is that the minipage environment can contain displayed text environments discussed in Chapter 6. The minipage environment can also contain footnotes (see Section 5.5.2) that are displayed within the minipage. See Section 3.2.1 of The LATEX Companion, 2nd edition [46] for complications that may arise therefrom.
5.9.4
Marginal comments A variant of the paragraph box, the \marginpar command, allows you to add marginal comments. So Do not use this \marginpar{Do not use this much.} much. produces the comment displayed in the margin. The AMS warning in the book [30] (also displayed here below the marginal comment) is defined as \marginpar{{\Large% \textcircled{\raisebox{.7pt}{\normalsize\textbf A}}}} The \textcircled command is discussed in Section 5.4.7, while the \raisebox command is introduced in Section 5.9.6.
Rule
Tip
Marginal comments and math environments Do not use marginal comments in equations or multiline math environments.
Avoid using too many marginal comments on any given page—LATEX may have to place some of them on the next page. If the document is typeset twosided, then the marginal comments are set in the outside margin. The form \marginpar[leftcomment ]{rightcomment } uses the required argument rightcomment when the marginal comment is set in the right margin and the optional argument leftcomment when the marginal comment is set in the left margin.
A
5.9
113
Boxes
The width of the paragraph box for marginal comments is stored in the length command \marginparwidth (see Section 15.5.2 for length commands). If you want to change it, use \setlength{\marginparwidth}{new_width } as in \setlength{\marginparwidth}{90pt} The default value of this width is set by the document class. If you want to know the present setting, type \the\marginparwidth in your document and typeset it, or, in interactive mode (see Sections 15.1.7 and D.4), type *\showthe\marginparwidth (* is the interactive prompt). See Sections 3.2.8 and 4.1 of The LATEX Companion, 2nd edition [46] for other style parameters pertaining to marginal notes.
5.9.5 Solid boxes A solid filled box is created with a \rule command. The first argument is the width and the second is the height. For instance, to obtain end of proof symbol: type end of proof symbol: \rule{1.6ex}{1.6ex} In fact, this symbol is usually slightly lowered: end of proof symbol: This positioning is done with an optional first argument: end of proof symbol: \rule[.23ex]{1.6ex}{1.6ex} Here is an example combining \rule with \makebox and \hrulefill: 1 inch:\quad\makebox[1in]{\rule{.4pt}{4pt}% \hrulefill\rule{.4pt}{4pt}}
114
Chapter 5
Typing text
which produces 1 inch:
Struts Solid boxes of zero width are called struts. Struts are invisible, but they force LATEX to make room for them, changing the vertical alignment of lines. Standard struts can also be added with the \strut or \mathstrut command. To see how struts work, compare ab and ab and ab typed as \fbox{ab} and \fbox{\strut ab}
and \fbox{$\mathstrut$ab}
Struts are especially useful for fine tuning tables (see Section 6.6, page 139) and formulas (see math struts in Section 8.4).
Rule
Zero distance 0pt, 0in, 0cm, 0em all stand for zero width. 0 by itself is not acceptable. For example, \rule{0}{1.6ex} gives the error message ! Illegal unit of measure (pt inserted). h l.251 \rule{0}{1.6ex} If the \rule command has no argument or only one, LATEX generates an error message. For instance, \rule{1.6ex} gives the message ! Paragraph ended before \@rule was complete. or ! Missing number, treated as zero. In the first error message, the reference to \@rule suggests that the problem is with the \rule command. Checking the syntax of the \rule command, you find that an argument is missing. The second error message is more informative, since there is, indeed, a missing number.
5.9
115
Boxes
5.9.6 Fine tuning boxes The command \raisebox{displacement }{text } typesets text in a box with a vertical displacement. If displacement is positive, the box is raised; if it is negative, the box is lowered. The \raisebox command allows us to play games: fine\raisebox{.5ex}{tun}\raisebox{.5ex}{ing} produces finetuning.
The \raisebox command has two optional arguments:
\raisebox{0ex}[1.5ex][0.75ex]{text } forces LATEX to typeset text as if it extended 1.5 ex above and 0.75 ex below the line, resulting in a change in the interline space above and below the line. A simple version of this command, \smash, is discussed in Section 8.4. In the AMS warning in the book [30] (shown on page 112), the \raisebox command is used to properly center the bold A in the circle: \Large\textcircled{\raisebox{.7pt}{\normalsize\textbf A}}
CHAPTER
6 Text environments
There are three types of text environments in LATEX: 1. Displayed text environments; text within such an environment usually is typeset with some vertical space around it 2. Text environments that create a “large symbol” 3. Style and size environments We start by discussing a very important rule about blank lines in displayed text environments. Then we proceed in Section 6.2 to the most often used displayed text environments: lists. We continue with the style and size environments in Section 6.3. The most important displayed text environments in math are proclamations or theoremlike structures, proclamations with style, and the proof environment, discussed in detail in Sections 6.4 and 6.5. The tabular environment discussed in Section 6.6 produces a “large symbol”, a table, which is of limited use in math.
118
Chapter 6
Text environments
In Section 6.7, we discuss the tabbing environment, which is often used for computer code. The legacy environments quote, quotation, and verse are discussed in Section 6.8, along with the verbatim environment, which is often used to display LATEX source in a typeset LATEX document.
6.1
Some general rules for displayed text environments As you know, blank lines play a special role in LATEX, usually indicating a paragraph break. Since displayed text environments structure the printed display themselves, the rules about blank lines are relaxed somewhat. However, a blank line trailing an environment signifies a new paragraph for the text following the environment.
Rule
Blank lines in displayed text environments 1. Blank lines are ignored immediately after \begin{name } or immediately before \end{name } except in a verbatim environment. 2. A blank line after \end{name } forces the text that follows to start a new paragraph. 3. As a rule, you should not have a blank line before \begin{name }. 4. The line after any theorem or proof always begins a new paragraph, even if there is no blank line or \par command. The page breaking commands in Section 5.7.3 apply to text environments, as does the line breaking command \\ discussed in Section 5.7.1.
6.2
List environments LATEX provides three list environments: enumerate, itemize, and description. LATEX also provides a generic list environment that can be customized to fit your needs. See Section 15.6 on custom lists. Most document classes redefine the spacing and some stylistic details of lists, especially since the list environments in the legacy document classes are not very pleasing. In this section, the list environments are formatted as they are by our standard document class, amsart. Throughout the rest of the book, lists are formatted as specified by this book’s designer.
6.2
List environments
119
6.2.1 Numbered lists A numbered list is created with the enumerate environment:
typed as \noindent This space has the following properties: \begin{enumerate} \item Grade 2 Cantor\label{Cantor}; \item Halfsmooth Hausdorff\label{Hausdorff}; \item Metrizably smooth\label{smooth}. \end{enumerate} Therefore, we can apply the Main Theorem. Each item is introduced with an \item command. The numbers LATEX generates can be labeled and crossreferenced (see Section 10.4.2). This construct can be used in theorems and definitions, for listing conditions or conclusions. If you use \item in the form \item[], you get an unnumbered item in the list.
6.2.2 Bulleted lists A bulleted list is created with the itemize environment:
is typed as \noindent We set out to accomplish a variety of goals: \begin{itemize} \item To introduce the concept of smooth functions. \item To show their usefulness in differentiation. \item To point out the efficacy of using smooth functions in Calculus. \end{itemize}
120
6.2.3
Chapter 6
Text environments
Captioned lists In a captioned list each item has a title (caption) specified by the optional argument of the \item command. Such lists are created with the description environment:
is typed as \noindent In this introduction, we describe the basic techniques: \begin{description} \item[Chopped lattice] a reduced form of a lattice; \item[Boolean triples] a powerful lattice construction; \item[Cubic extensions] a subdirect power flattening the congruences. \end{description}
6.2.4
A rule and combinations There is only one rule you must remember.
Rule
List environments An \item command must immediately follow \begin{enumerate}, \begin{itemize}, or \begin{description}. Of course, spaces and line breaks can separate them. If you break this rule, you get an error message. For instance, \begin{description} This is wrong! \item[Chopped lattice] a reduced lattice; gives the error message ! LaTeX Error: Something’s wrongperhaps a missing \item. l.105
\item[Chopped lattice] a reduced lattice;
6.2
List environments
121
If you see this error message, remember the rule for list environments and check for text preceding the first \item. You can nest up to four list environments; for instance,
which is typed as \begin{enumerate} \item First item of Level 1. \begin{enumerate} \item First item of Level 2. \begin{enumerate} \item First item of Level 3. \begin{enumerate} \item First item of Level 4. \item Second item of Level 4.\label{level4} \end{enumerate} \item Second item of Level 3. \end{enumerate} \item Second item of Level 2. \end{enumerate} \item Second item of Level 1. \end{enumerate} Referencing the second item of Level 4: \ref{level4} Note that the label level4 collected all four of the counters (see Section 10.4.2). You can also mix list environments:
122
Chapter 6
Text environments
which is typed as \begin{enumerate} \item First item of Level 1. \begin{itemize} \item First item of Level 2. \begin{enumerate} \item First item of Level 3. \begin{itemize} \item First item of Level 4. \item Second item of Level 4.\label{enums} \end{itemize} \item Second item of Level 3. \end{enumerate} \item Second item of Level 2. \end{itemize} \item Second item of Level 1. \end{enumerate} Referencing the second item of Level 4: \ref{enums} Now the label enums collects only the two enumerate counters (see Section 10.4.2). The indentations are, of course, not needed. I use them to keep track of the level of nesting. In all three types of list environment, the \item command may be followed by an optional argument, which is displayed at the beginning of the typeset item: \item[label ] Note that for enumerate and itemize the resulting typography may leave something to be desired.
6.3
Style and size environments
123
Tip
If the text following an \item command starts with an opening square bracket, [, then LATEX thinks that \item has an optional argument. To prevent this problem from occurring, type [ as {[}. Similarly, a closing square bracket, ], inside the optional argument should be typed as {]}.
Tip
You may want to use a list environment solely for the way the items are displayed, without any labels. You can achieve this effect by using \item[]. You can change the style of the numbers in an enumerate environment by redefining the counter as suggested in Section 15.5.1: \renewcommand{\labelenumi}{{\normalfont (\roman{enumi})}} The labels then are displayed as (i), (ii), and so on. This modification only works if you do not want to reference these items. If you want the \ref command to work properly, use David Carlisle’s enumerate package (see Section 12.3.1). For an example of how to use Carlisle’s environment, see Section 15.2.1. Section 3.3 of The LATEX Companion, 2nd edition [46] explains how to customize the three list environments and discusses Bernd Schandl’s paralist package, which provides a number of new list environments and makes customizing the three legacy list environments much easier. For custom lists, see Section 15.6.
6.3
Style and size environments There are several text environments that allow you to set font characteristics. They have the same names as their corresponding command declarations: rmfamily sffamily ttfamily upshape itshape em slshape scshape bfseries For instance, \begin{ttfamily} text \end{ttfamily} typesets text just like {\ttfamily text } would. Remember to use the commanddeclaration names for the environment names, that is, use rmfamily, not textrm and ttfamily, not texttt (see Section 5.6.2). There are also text environments for changing the font size, from tiny to Huge (see Section 5.6.6).
124
Chapter 6
Text environments
If you are getting overwhelmed by the large number of environments changing style and size, consult Tables 5.6 and 5.7 (see also Section C.3.2). Horizontal alignment of a paragraph is controlled by the flushleft, flushright, and center environments. Within the flushright and center environments, it is customary to force new lines with the \\ command, while in the flushleft environment, you normally allow LATEX to wrap the lines. These text environments can be used separately or in combination, as in The simplest text environments set the printing style and size. The commands and the environments have similar names. typed as \begin{flushright} The \begin{bfseries}simplest\end{bfseries} text environments set the printing style and size.\\ The commands and the environments have similar names. \end{flushright} There are command declarations that correspond to these environments: \centering centers text \raggedright left aligns text \raggedleft right aligns text The effect of one of these commands is almost the same as that of the corresponding environment except that the environment places additional vertical space before and after the displayed paragraphs. For such a command declaration to affect the way a paragraph is formatted, the scope must include the whole paragraph, including the blank line at the end of the paragraph, preferably indicated with a \par command. The \centering command is used often with the \includegraphics command (see Section 10.4.3).
6.4
Proclamations (theoremlike structures) Theorems, lemmas, definitions, and so forth are a major part of mathematical writing. In LATEX, these constructs are typed in displayed text environments called proclamations or theoremlike structures. In the intrart.tex sample article (see pp. 42–43), there are two theorems, a definition, and a notation. These four environments have similar structures, only their names are different.
6.4
Proclamations (theoremlike structures)
125
In the sampart.tex sample article (see pp. 286–288), there are a number of different proclamations in a variety of styles, with varying degrees of emphasis. Proclamations with style are discussed in Section 6.4.2. We discuss in Section 4.2.3 the two steps required by proclamations: Step 1 Define the proclamation with a \newtheorem command in the preamble of the document. For instance, the line \newtheorem{theorem}{Theorem} defines a theorem environment. Step 2 Invoke the proclamation as an environment in the body of your document. Using the proclamation definition from Step 1, type \begin{theorem} My first theorem. \end{theorem} to produce a theorem:
Theorem 1. My first theorem. In the proclamation definition \newtheorem{theorem}{Theorem} the first argument, theorem, is the name of the environment that invokes the theorem. The second argument, Theorem, is the name that is used when the proclamation is typeset. LATEX numbers the theorems automatically and typesets them with vertical space above and below. The phrase Theorem 1. appears, followed by the theorem itself, which may be emphasized. Of course, the formatting of the theorem depends on the document class and on the proclamation style (see Section 6.4.2). You may also specify an optional argument, \begin{theorem}[The FuchsSchmidt Theorem] The statement of the theorem. \end{theorem} that appears as the name of the theorem:
Theorem 1 (The FuchsSchmidt Theorem). The statement of the theorem.
126
Chapter 6
Text environments
LATEX is very fussy about how proclamations are defined. For example, in the introductory article intrart.tex (see Section 4.1), if the closing brace is dropped from the end of line 8, \newtheorem{definition}{Definition you get an error message such as Runaway argument? {Definition \newtheorem {notation}{Notation} ! Paragraph ended before \@ynthm was complete. \par l.10 Line 10 is the line after the \newtheorem commands. The message conveys the information that something is wrong in the paragraph before line 10. If you forget an argument, as in \newtheorem{definition} LATEX produces an error message such as ! LaTeX Error: Missing \begin{document}. l.9 \newtheorem{n
otation}{Notation}
In the error message, the line ! LaTeX Error: Missing \begin{document}. usually means that LATEX became confused and believes that some text typed in the preamble should be moved past the line \begin{document} The mistake could be anywhere in the preamble above the line LATEX indicates. If you encounter such an error message, try to isolate the problem by commenting out parts of the preamble (see Section 5.5.1 and also Section 4.3.1).
Rule
Lists in proclamations If a proclamation starts with a list environment, precede the list by \hfill.
6.4
Proclamations (theoremlike structures)
127
If you do not, as in \begin{definition}\label{D:prime} \begin{enumerate} \item $u$ is \emph{bold} if $u = x^2$.\label{mi1} \item $u$ is \emph{thin} if $u = \sqrt{x}$.\label{mi2} \end{enumerate} \end{definition} your typeset list starts on the first line of the proclamation:
If you add the \hfill command, \begin{definition}\hfill \begin{enumerate} the list in the definition typesets correctly:
Consecutive numbering If you want to number two sets of proclamations consecutively, you can do so by first defining one proclamation, and then using its name as an optional argument of the second proclamation. For example, to number the lemmas and propositions in your paper consecutively, you type the following two lines in your preamble: \newtheorem{lemma}{Lemma} \newtheorem{proposition}[lemma]{Proposition} Lemmas and propositions are then consecutively numbered as Lemma 1, Proposition 2, Proposition 3, and so on. Let me emphasize: The optional argument of a proclamation definition must be the name of a proclamation that has already been defined. Numbering within a section The \newtheorem command may also have a different optional argument; it causes LATEX to number the lemmas within sections. For example,
128
Chapter 6
Text environments
\newtheorem{lemma}{Lemma}[section] numbers the lemmas in Section 1 as Lemma 1.1 and Lemma 1.2. In Section 2, you have Lemma 2.1 and Lemma 2.2, and so on. Instead of section, you may use any sectioning command provided by the document class, such as chapter, section, and subsection. Consecutive numbering and numbering within a section can be combined. For example, \newtheorem{lemma}{Lemma}[section] \newtheorem{proposition}[lemma]{Proposition} sets up the lemma and proposition environments so that they are numbered consecutively within sections: Lemma 1.1, Proposition 1.2, Proposition 1.3 and Proposition 2.1, Lemma 2.2, and so on.
6.4.1
The full syntax The full form of \newtheorem is \newtheorem{envname }[procCounter ]{Name }[secCounter ] where the two optional arguments are mutually exclusive, and envname is the name of the environment to be used in the body of the document. For instance, you may use theorem for the envname of a theorem, so that a theorem is typed inside a theorem environment. Of course, envname is just a label; you are free to choose any environment name, such as thm or george (as long as the name is not in use as the name of another command or environment). This argument is also the name of the counter LATEX uses to number these text environments. procCounter is an optional argument. It sets the new proclamation to use the counter of a previously defined proclamation and the two proclamations are consecutively numbered. Name is the text that is typeset when the proclamation is invoked. So if Theorem is given as Name, then you get Theorem 1, Theorem 2, and so on in your document. secCounter is an optional argument that causes Name environments to be numbered within the appropriate sectioning units. So if theorem is the envname and section is the secCounter, then in Section 1 you have Theorem 1.1, Theorem 1.2, and so on. In Section 2 you get Theorem 2.1, Theorem 2.2, and so on. Proclamations may be numbered within subsections, sections, chapters, or any other sectioning unit automatically numbered by LATEX.
6.4
Proclamations (theoremlike structures)
129
6.4.2 Proclamations with style You can choose one of three styles for your proclamations by preceding the definitions with the \theoremstyle{style} command, where style is one of the following: plain, the most emphatic definition remark, the least emphatic There are a few extra options, including the \newtheorem* command, an unnumbered version of \newtheorem. The following commands set the styles in the sampart.tex article (see page 290). The typeset sample article (on pages 286–288) shows how the chosen styles affect the typeset proclamations. \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{corollary}{Corollary} \newtheorem*{main}{Main Theorem} \newtheorem{lemma}{Lemma} \newtheorem{proposition}{Proposition} \theoremstyle{definition} \newtheorem{definition}{Definition} \theoremstyle{remark} \newtheorem*{notation}{Notation} A proclamation created by a \newtheorem command has the style of the last \theoremstyle command preceding it. The default style is plain. Three examples Here are three sets of proclamation definitions to illustrate different styles and numbering schemes. Example 1 \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{lemma}[theorem]{Lemma} \newtheorem{definition}[theorem]{Definition} \newtheorem{corollary}[theorem]{Corollary}
130
Chapter 6
Text environments
In a document with this set of proclamation definitions you can use theorems, lemmas, definitions, and corollaries, typeset in the most emphatic (plain) style. They are all numbered consecutively: Definition 1, Definition 2, Theorem 3, Corollary 4, Lemma 5, Lemma 6, Theorem 7, and so on. Example 2 \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem*{main}{Main Theorem} \newtheorem{definition}{Definition}[section] \newtheorem{lemma}[definition]{Lemma} \theoremstyle{definition} \newtheorem*{Rule}{Rule} In this document you may use theorems, definitions, and lemmas in the most emphatic (plain) style, and unnumbered rules in the less emphatic (definition) style. Definitions and lemmas are numbered consecutively within sections. You may also use the unnumbered Main Theorem. So, for example, you may have Definition 1.1, Definition 1.2, Main Theorem, Rule, Lemma 1.3, Lemma 2.1, Theorem 1, and so on. Example 3 \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{corollary}{Corollary} \newtheorem*{main}{Main Theorem} \newtheorem{lemma}{Lemma} \newtheorem{proposition}{Proposition} \theoremstyle{definition} \newtheorem{definition}{Definition} \theoremstyle{remark} \newtheorem*{notation}{Notation} With these proclamation definitions you can use theorems, corollaries, lemmas, and propositions in the most emphatic (plain) style, and an unnumbered Main Theorem. You can have definitions in the less emphatic (definition) style. All are separately numbered. So in the document you may have Definition 1, Definition 2, Main Theorem, Lemma 1, Proposition 1, Lemma 2, Theorem 1, Corollary 1, and so on. You can also have Notations which are unnumbered and typeset in the least emphatic (remark) style.
6.5
Proof environments
131
Number swapping Proclamations can be numbered on the left, as for instance, 3.2 Theorem. To accomplish this, type the \swapnumbers command before the \newtheorem command corresponding to the proclamation definition you want to change. This command affects all of the proclamation definitions that follow it, so the proclamation definitions in the preamble should be in two groups. The regular ones should be listed first, followed by the \swapnumbers command, then all the proclamations that swap numbers. Do not swap numbers unless the publisher demands it. Custom theorem styles You can define custom theorem styles with the \newtheoremstyle command. You should very seldom do this, the three theorem styles of the document class should suffice. For more detail, see [5].
6.5
Proof environments A proof is the contents of a proof environment. For instance, Proof. This is a proof, delimited by the q.e.d. symbol. typed as \begin{proof} This is a proof, delimited by the q.e.d.\ symbol. \end{proof} A proof is set off from the surrounding text with some vertical space. The end of the proof is marked with the symbol at the end of the line. There are a few examples of the proof environment in the sampart.tex sample article (pages 286–293). We start with the same rule for proofs as we have for proclamations on page 126.
Rule
Lists in proofs If a proof starts with a list environment, precede the list by \hfill. If you want to suppress the symbol at the end of a proof, give the command \begin{proof} ... \renewcommand{\qedsymbol}{} \end{proof}
132
Chapter 6
Text environments
To suppress the end of the proof symbol in the whole article, give the \renewcommand{\qedsymbol}{} command in the preamble. To substitute another phrase for Proof, such as Necessity, as in Necessity. This is the proof of necessity. use the proof environment with an optional argument: \begin{proof}[Necessity] This is the proof of necessity. \end{proof} The optional argument may contain a reference, as in \begin{proof}[Proof of Theorem~\ref{T:smooth}] which might be typeset as Proof of Theorem 5. This is the proof. It is easy to make the mistake of placing the optional argument after \begin: \begin[Proof of Theorem~\ref{T:P*}]{proof} You get an error message ! LaTeX Error: Bad math environment delimiter. l.91 \begin{equation} \label{E:cong2} which is not very helpful. There is a problem with the placement of the q.e.d. symbol if the proof ends with a displayed formula (or a list environment). For instance, \begin{proof} Now the proof follows from the equation \[ a^2 = b^2 + c^2. \] \end{proof}
6.6
133
Tabular environments typesets as Proof. Now the proof follows from the equation a2 = b2 + c2 .
To correct the placement of the q.e.d. symbol, use the \qedhere command: \begin{proof} Now the proof follows from the equation \[ a^2 = b^2 + c^2.\qedhere \] \end{proof} which typesets as Proof. Now the proof follows from the equation a2 = b2 + c2 .
6.6
Tabular environments A tabular environment creates a table that LATEX treats as a “large symbol”. In particular, a table cannot be broken across pages. Name 1 2 3 Peter 2.45 34.12 1.00 Here is a simple table, , typeset inline. This John 0.00 12.89 3.71 David 2.00 1.85 0.71 looks awful, but it does make the point that the table is just a “large symbol”. The table is typed as \begin{tabular}{  l  r  r  r  } \hline Name & 1 & 2 & 3 \\ Peter & 2.45 & 34.12 & 1.00\\ John & 0.00 & 12.89 & 3.71\\ David & 2.00 & 1.85 & 0.71\\ \end{tabular}
\hline \hline \hline \hline
with no blank line before or after the environment.
134
Chapter 6
Text environments
This table can be horizontally centered with a center environment (see Section 6.3). It can also be placed within a table environment (see Section 10.4.3). This sets the table off from the surrounding text with vertical space and you can also use the float controls b, t, h, p to specify where the table should appear (see Section 10.4.3). This also allows you to define a caption, which can be placed before or after the table: \begin{table} \begin{center} \begin{tabular}{  l  r  r  r  } \hline Name & 1 & 2 & 3 \\ \hline Peter & 2.45 & 34.12 & 1.00\\ \hline John & 0.00 & 12.89 & 3.71\\ \hline David & 2.00 & 1.85 & 0.71\\ \hline \end{tabular} \caption{Tabular table.}\label{Ta:first} \end{center} \end{table} This table is displayed as Table 6.1. Name Peter John David
1 2.45 0.00 2.00
2 34.12 12.89 1.85
3 1.00 3.71 0.71
Table 6.1: Tabular table. It can be listed in a list of tables (see Section 10.4.3) and the table number may be referenced using the command \ref{Ta:first}. Note that the label must be typed between the caption and the \end{table} command. For another example, look at the two tables in the fonttbl.tex file in your samples folder. The first is typed as \begin{tabular}{rllllllllll} & 0 & 1 & 2 & 3 & 4 & 5 & 6 & 7 & 8 & 9\\ \hline 0& \symbol{0} &\symbol{1}&\symbol{2}&\symbol{3}& \symbol{4}&\symbol{5}&\symbol{6}&\symbol{7}& \symbol{8}&\symbol{9}\\ \hline . . .
6.6
Tabular environments
135
120& \symbol{120} &\symbol{121}&\symbol{123}& \symbol{123}&\symbol{124}&\symbol{125}&\symbol{126} &\symbol{127} && \\ \hline \end{tabular} The second table is the same except that the numbers run from 128 to 255. The typeset table is shown in Section 5.4.4.
Rule
tabular environments 1. \begin{tabular} requires an argument consisting of a character l, r, or c, meaning left, right, or center alignment, for each column, and optionally, the  symbols. Each  indicates a vertical line in the typeset table. Spaces in the argument are ignored but can be used for readability. 2. Columns are separated by ampersands (&) and rows are separated by \\. 3. & absorbs spaces on either side. 4. The \hline command creates a horizontal rule in the typeset table. It is placed either at the beginning of the table (after the \begin line) or it must follow a \\ command. 5. If you use a horizontal line to finish the table, you must separate the last row of the table from the \hline command with the \\ command. 6. \begin{tabular} takes an optional argument, b or t, to specify the bottom or the top vertical alignment of the table with the baseline. The default is center alignment.
Remember to put the optional argument b or t in square brackets, as in \begin{tabular}[b]{  l  r  r  r  } If you forget to place an \hline command right after \\ in the last row, you get an error message such as ! Misplaced \noalign. \hline >\noalign {\ifnum 0=‘}\fi \hrule \@height \arrayrulew... l.9 ....00 & 1.85 & 0.71 \hline
136
Chapter 6
Text environments
More columnformatting commands The required argument of the tabular environment may contain columnformatting commands of various types. An @expression, for instance, @{.}, replaces the space LATEX normally inserts between two columns with its argument. For example, \begin{tabular}{r @{.} l} 3&78\\ 4&261\\ 4 \end{tabular} creates a table with two columns separated by a decimal point. In effect, you get a single, decimalaligned column: 3.78 4.261 4. This example is an illustration. You should use David Carlisle’s dcolumn package if you need a decimalaligned column (see Section 12.3.1). The width of a column depends on the entries in the column by default. You can specify a width by using the p column specifier: p{width } For instance, if you want the first column of Table 6.1 to be 1 inch wide, then type \begin{tabular}{  p{1in}  Name & 1 & 2 Peter & 2.45 & 34.12 John & 0.00 & 12.89 David & 2.00 & 1.85 \end{tabular}
r & & & &
 r  r  }\hline 3 \\ \hline 1.00\\ \hline 3.71\\ \hline 0.71\\ \hline
which typesets as Name Peter John David
1 2.45 0.00 2.00
2 34.12 12.89 1.85
3 1.00 3.71 0.71
6.6
137
Tabular environments
To center the items in the first column, precede each item with a \centering command (see Section 6.3). Note that the first column is actually somewhat over 1 inch wide, because of the extra space provided around the column boundaries. The p column specifier can also be used for multiline entries. Refinements \hline draws a horizontal line the whole width of the table. \cline{ab } draws a horizontal line from column a to column b. For instance, \cline{13} or
\cline{44}
Another useful command is \multicolumn, which is used to span more than one column, for example, \multicolumn{3}{c}{\emph{absent}} The first argument is the number of columns spanned by the entry, the second is the alignment (an optional vertical line designator  for this row only can also be included), and the third argument is the entry. Note that the entry for the spanned columns is in braces. An example is shown in Table 6.2, typed as follows: \begin{table}[h!] Name Peter John David
1 2.45
2 34.12 absent 2.00 1.85
3 1.00 0.71
Table 6.2: Table with \multicolumn. Name Peter
Month Jan.
John
Total Jan.
Total Grand Total
Week 1 2 3 4 1 2 3 4
Amount 1.00 12.78 0.71 15.00 29.49 12.01 3.10 10.10 0.00 25.21 54.70
Table 6.3: Table with \multicolumn and \cline.
138
Chapter 6
Text environments
\begin{center} \begin{tabular}{  l  r  r  r  } \hline Name & 1 & 2 & 3\\ \hline Peter & 2.45 & 34.12 & 1.00\\ \hline John & \multicolumn{3}{c }{\emph{absent}}\\ \hline David & 2.00 & 1.85 & 0.71\\ \hline \end{tabular} \caption{Table with \bsl\texttt{multicolumn}.} \label{Ta:mc} \end{center} \end{table} The next example, shown in Table 6.3, uses the \multicolumn and \cline commands together: \begin{table}[t] \begin{center} \begin{tabular}{  c c  c  r  } \hline Name & Month & Week & Amount\\ \hline Peter & Jan. & 1 & 1.00\\ \cline{34} & & 2 & 12.78\\ \cline{34} & & 3 & 0.71\\ \cline{34} & & 4 & 15.00\\ \cline{24} & \multicolumn{2}{ l}{Total} & 29.49\\ \hline John & Jan. & 1 & 12.01\\ \cline{34} & & 2 & 3.10\\ \cline{34} & & 3 & 10.10\\ \cline{34} & & 4 & 0.00\\ \cline{24} & \multicolumn{2}{ l}{Total} & 25.21\\ \hline \multicolumn{3}{l}{Grand Total} & 54.70\\ \hline \end{tabular} \caption{Table with \bsl\texttt{multicolumn} and \bsl\texttt{cline}.}\label{Ta:multicol+cline} \end{center} \end{table} The \parbox command (see Section 5.9.3) can be used to produce a single multiline entry. Recall that the first argument of \parbox is the width of the box. A p{} width designator creates a column in which all entries can be multiline. As an example, to replace Grand Total by Grand Total for Peter and John, type the last line as
6.6
139
Tabular environments \multicolumn{3}{l}{ \parbox[b]{10em}{Grand Total\\ for Peter and John} } & 54.70\\ \hline
Note the use of the bottom alignment option (see Section 5.9.3). The last row of the modified table prints Grand Total for Peter and John
54.70
The spacing above Grand Total is not quite right. It can be adjusted with a strut (see Section 5.9.5), \parbox[b]{10em}{\strut Grand Total\\ for Peter and John:} Finally, vertical spacing can be adjusted by redefining \arraystretch. For instance, in the table
5th Grade: 6th Grade: Overall:
Area 63.4 m2 62.0 m2 62.6 m2
Students 22 19 20
typed as \begin{center} \begin{tabular}{rcc}\hline & \textbf{Area} & \textbf{Students}\\ \hline \textbf{5th Grade}: & 63.4 m\textsuperscript{2} &22\\ \hline \textbf{6th Grade}: & 62.0 m\textsuperscript{2} &19\\ \hline \textbf{Overall}: & 62.6 m\textsuperscript{2} &20\\ \hline \end{tabular} \end{center} you may find that the rows are too crowded. The vertical spacing may be adjusted by adding the line \renewcommand{\arraystretch}{1.25} to the tabular environment. To limit its scope, add it after \begin{center}
140
Chapter 6
Text environments
The adjusted table is typeset as Area 5th Grade: 6th Grade: Overall:
Students
63.4 m
2
22
62.0 m
2
19
62.6 m2
20
In some tables, horizontal and vertical lines do not always intersect as desired. Fine control over these intersections is provided by the hhline package (see Section 12.3.1). Chapter 5 of The LATEX Companion, 2nd edition [46] deals with tabular material, discussing many extensions, including multipage tables, decimalpoint alignment, footnotes in tables, tables within tables, and so on.
6.6.1
Table styles LATEX can draw double horizontal and vertical lines in tables with ease. As a result, there are far too many double lines in LATEX tables, resulting in cluttered and confusing tables. The Chicago Manual of Style, 15th edition [11] has almost 80 pages on tables. For simple tables it advocates a simple style, as shown in Table 6.4. Notice the generous space above and below the column heads, which has been achieved with the \rule[8pt]{0pt}{22pt} command, some extra space above the first line of data, which has been achieved with the \rule{0pt}{14pt} command, the columns of equal width, which has been achieved with p{70pt} commands, and no vertical lines. Most tables in this book have been designed according to this style using Simon Fear’s booktabs package (see Section E.1). Table 6.4: Smokers and Nonsmokers, by Sex. Smoke
Don’t Smoke
Total
Males
1,258
2,104
3,362
Females
1,194
2,752
3.946
2,452
4,856
7,308
Total
6.7
6.7
Tabbing environments
141
Tabbing environments Although of limited use for mathematical typesetting, the tabbing environment can be useful for typing algorithms, computer programs, and so forth. LATEX calculates the width of a column in the tabular environment based on the widest entry (see Section 6.6). The tabbing environment allows you to control the width of the columns. The \\ command is the line separator, tab stops are set by \= and are remembered by LATEX in the order they are given, and \> moves to the next tab position. You can easily reset tab positions. For instance, if you are past the second tab position by using \> twice, and there is a third tab position, the \= command resets it. Lines of comments may be inserted with the \kill command, see the examples below, or with the % character. The difference is that a line with \kill can be used to set tab stops, whereas a commented out line cannot. A simple example:
PrintTime Block[timing], timing = Timing[expr]; Print[ timing[[1]] ]; ] End[ ]
typed as {\ttfamily \begin{tabbing} Print\=Time\\ \>Block\=[{timing},\\ \>\>timing = Timing[expr];\\ (careful with initialization)\kill \>\>Print[ timing[[1]] ];\\ \>]\\ End[] \end{tabbing} }% end \ttfamily An alternative method is to use a line to set the tab stops, and then \kill the line so it does not print: {\ttfamily \begin{tabbing} \hspace*{.25in}\=\hspace{2ex}\=\hspace{2ex}\=
142
Chapter 6
Text environments
\hspace{2ex}\kill \> $k := 1$\\ \> $l_k := 0$; $r_k := 1$\\ \> loop\\ \> \> $m_k := (l_k + r_k)/2$\\ \> \> if $w < m_k$ then\\ \> \> \> $b_k := 0$; $r_k := m_k$\\ \> \> else if $w > m_k$ then\\ \> \> \> $b_k := 1$; $l_k := m_k$\\ \> \> end if\\ \> \> $k := k + 1$\\ \> end loop \end{tabbing} }% end \ttfamily which typesets as
k := 1 lk := 0; rk := 1 loop mk := (lk + rk )/2 if w < mk then bk := 0; rk := mk else if w > mk then bk := 1; lk := mk end if k := k + 1 end loop
Some simple rules: There is no \\ command on a line containing the \kill command. You may set the tabs in a \kill line with \hspace commands. The \> command moves to the next tab stop, even if the text you have already typed extends past that stop, which can result in overprinting. The tabbing environment has to be typeset with typewriter style font—note the \ttfamily command.
6.8
Miscellaneous displayed text environments
143
To illustrate the third rule, type \begin{tabbing} This is short.\=\\ This is much longer, \> and jumps back. \end{tabbing} which typesets as This is short. This is muchand longer, jumps back. If you do not follow the fourth rule, be careful with your tabbing. You do not really have to use typewriter style font—just beware of the pitfalls. For more information, please consult Chapter 5 of The LATEX Companion, 2nd edition [46].
6.8
Miscellaneous displayed text environments There are four more displayed text environments, of limited use in math: quote, quotation, verse, and verbatim. We also discuss an inline version of the verbatim environment, the \verb command. Quotes The quote environment is used for short (one paragraph) quotations:
which is typed as: \begin{quote} It’s not that I’m afraid to die. I just don’t want to be there when it happens. \emph{Woody Allen} Literature is news that STAYS news. \emph{Ezra Pound} \end{quote} Note that multiple quotes are separated by blank lines.
144
Chapter 6
Text environments
Quotations In the quotation environment, blank lines mark new paragraphs:
is typed as \begin{quotation} KATH: Can he be present at the birth of his child? ED: It’s all any reasonable child can expect if the dad is present at the conception. \begin{flushright} \emph{Joe Orton} \end{flushright} \end{quotation} Verses A verse environment,
is typed as \begin{verse} I think that I shall never see\\ A poem lovely as a tree. Poems are made by fools like me,\\ But only God can make a tree. \begin{flushright} \emph{Joyce Kilmer}
6.8
Miscellaneous displayed text environments
145
\end{flushright} \end{verse} Lines are separated by \\ and stanzas by blank lines. Long lines are typeset with hanging indent. Verbatim typesetting Finally, there is the verbatim text environment. You may need it if you write about LATEX or some other computer program or if you have to include portions of a source file or user input in your typeset work. Most of the displayed source in this book was written in a verbatim environment. For instance, you may have to write to a journal about an article you are proofreading: Formula (2) in Section 3 should be typed as follows: \begin{equation} D = \{\, x_0 \mid x_0 \Rightarrow a_1 \,\} \tag{2} \end{equation} Please make the necessary corrections. The problem is that if you just type Formula (2) in Section 3 should be typed as follows: \begin{equation} D = \{\, x_0 \mid x_0 \Rightarrow a_1 \,\} \tag{2} \end{equation} Please make the necessary corrections. it typesets as Formula (2) in Section 3 should be typed as follows: (2)
D = { x0  x0 ⇒ a1 }
Please make the necessary corrections. To get the proper typeset form, type it as follows: Formula (2) in Section 3 should be typed as follows: \begin{verbatim} \begin{equation} D = \{\, x_0 \mid x_0 \Rightarrow a_1 \,\} \tag{2} \end{equation}
146
Chapter 6
Text environments
\end{verbatim} Please make the necessary corrections.
Rule
verbatim text environments A verbatim environment cannot be placed within Another verbatim environment The argument of a command The closing line, \end{verbatim}, must be on a line by itself. A violation of the first rule results in unmatched environment delimiters. You get an error message such as ! \begin{document} ended by \end{verbatim}. A violation of the second rule gives an error message such as ! Argument of \@xverbatim has an extra }.
Tip
There are two traps to avoid when using the verbatim environment. 1. If the \end{verbatim} line starts with spaces, a blank line is added to the typeset version. 2. Any characters following \end{verbatim} on the same line are dropped and you get a LATEX warning. To illustrate the first trap, type the last two lines of the previous example as follows: \end{verbatim} Please make the necessary corrections. Then you find an unintended blank line before the last line. The second trap can be seen if you type the last line of the above example as \end{verbatim} Please make the necessary corrections. When typeset, Please make the necessary corrections. does not appear, and you receive a warning LaTeX Warning: Characters dropped after ‘\end{verbatim}’ on input line 17.
6.8
147
Miscellaneous displayed text environments
Several improved versions of the verbatim environment are provided by the verbatim package (see Section 12.3.1). To use this package, include the command \usepackage{verbatim} in the preamble. In fact, the rules discussed in this section are those of the verbatim package. The verbatim environment has some interesting variants and a number of them are discussed in Section 3.4 of The LATEX Companion, 2nd edition [46]. For instance, the alltt package, which is part of the standard LATEX distribution (see Section 12.3) is used to type the command syntax in this book. See the full syntax of \newtheorem on page 128 for an example. Verbatim typesetting inline The verbatim environment also has an inline version called \verb. Here is an example: Some European email addresses contain \%; recall that you have to type \verb+\%+ to get \%. which prints Some European email addresses contain %; recall that you have to type \% to get %. The character following the \verb command is a delimiter. In this example I have used +. The argument starts with the character following the delimiter, and it is terminated by the next occurrence of the delimiter. In this example, the argument is \%. Choose the delimiter character carefully. For instance, if you want to typeset $\sin(\pi/2 + \alpha)$ verbatim, and you type \verb+$\sin(\pi/2 + \alpha)$+ then you get the error message ! Missing $ inserted. $ l.5 \verb+$\sin(\pi/2 + \alpha )$+ Indeed, the argument of \verb is $\sin(\pi/2 because the second + terminates the \verb command. Then LATEX tries to typeset \alpha)$+, but cannot because it is not in math mode. Use another character, such as !, in place of +: \verb!$\sin(\pi/2 + \alpha)$!
148
Rule
Chapter 6
Text environments
verb commands The entire \verb command must be on a single line of your source file. There can be no space between the \verb command and the delimiter. The \verb command cannot appear in the argument of another command. The \verb command cannot be used within an aligned math environment. Do not use * as a delimiter. If you violate the first rule, as in \verb!$\sin(\pi/2 + \alpha)$! you get the error message ! LaTeX Error: \verb command ended by end of line. l.6 \verb!$\sin(\pi/2 + The \verb command has a *ed version which prints spaces as symbols. For example, \today the is typed as \verb*+\today the+. The \verb command can perform the function of the verbatim environment. The last error message, which was displayed in a verbatim environment, may be typed as follows: you get the error message\\[8pt] \verb! LaTeX Error: \verb command ended by end of line.\\ \verb \\ \verbl.6 \verb!$\sin(\pi/2 +\\[8pt]
Rule
Simulating verbatim with verb 1. End the line before the verbatim environment with \\[8pt]. 2. Each line xxx of the verbatim environment is placed in the construct: \verbxxx  If  occurs in xxx, then choose a different delimiter.
6.8
Miscellaneous displayed text environments 3. The last line yyy of the verbatim environment is placed in the construct: \verbyyy \\[8pt] If  occurs in yyy, then choose a different delimiter.
149
CHAPTER
7 Typing math
LATEX was designed for typesetting math. I address this topic in detail. A math formula can be typeset inline, as part of the current paragraph, or displayed, on a separate line or lines with vertical space before and after the formula. In this and the next chapter we discuss formulas that are set inline or displayed on a single line. In Chapter 9 we address multiline math formulas. We start with a discussion of LATEX’s basic math environments (Section 7.1), spacing rules in math (Section 7.2), and continue with the equation environment (Section 7.3). The basic constructs of a formula—arithmetic (including subscripts and superscripts), binomial coefficients, ellipses, integrals, roots, and text—are discussed in detail in Section 7.4. From the basic constructs of that section, you can build very complicated formulas, one step at a time. The process is illustrated in Section 7.4.7. Delimiters, operators, and math accents are dealt with in Sections 7.5–7.7. In Section 7.8, we discuss three types of stretchable horizontal lines that can be used above or below a formula: braces, bars, and arrows. There are also stretchable arrow math symbols. Section 7.9 is our Formula Gallery, in which you find a large number of illustrations, some straightforward, some more imaginative, of the math constructs introduced in the preceding sections.
152
7.1
Chapter 7
Typing math
Math environments A formula inR a LATEX document can be typeset inline, like the congruence a ≡ b (θ) or √ 2 ∞ the integral −∞ e−x dx = π, or displayed, as in a ≡ b (θ) or
Z
∞
−∞
e−x dx = 2
√
π
Notice how changing these two formulas from inline to displayed affects their appearance. Inline and displayed math formulas are typeset using the math environments math and displaymath, respectively. Because math formulas occur so frequently, LATEX has abbreviations: the special braces \( and \) or $ are used for the math environment, and \[ and \] for the displaymath environment.1 So our inline example may be typed as $a \equiv b \pod{\theta}$ or \( a \equiv b \pod{\theta} \) or \begin{math} a \equiv b \pod{\theta} \end{math} The displayed example can be typed as \[
\int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi}
\] or \begin{displaymath} \int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi} \end{displaymath} Using $ as a delimiter for a math environment is a bit of an anomaly, since the same character is used as both an opening and closing delimiter. This dual purpose use makes it more difficult for LATEX to diagnose an incorrect use of $. For instance, 1 T X uses $$ to open and close a displayed math environment. In LAT X, this may occasionally cause E E problems. Don’t do it! Try the fleqn document class option of amsart (see Section 11.5) as an example of what can go wrong.
7.1
153
Math environments Let $a be a real number, and let $f$ be a function. would be interpreted by LATEX as follows: Let
is ordinary text
$a be a real number, and let $ f
is math
is interpreted as ordinary text
$ be a function.
is thought to be a math environment (opened by $) that should be closed by the next $ in the paragraph
Because the paragraph ends with no more dollar signs appearing, you get the error message !! Missing $ inserted. $ l.29 and giving you the line number of the end of the paragraph. This message tells you that LATEX would place a $ at the end of the paragraph when it proceeds with the typesetting. Press Return and LATEX produces the following: Let abearealnumber, andletfbeaf unction. The text that ended up in a math environment is run together because math environments ignore spaces (see Section 7.2). If you use \( and \) as special braces for the math environment, LATEX handles the same mistake more elegantly: Let \( a be a real number, and let \( f \) be a function. gives the error message ! LaTeX Error: Bad math environment delimiter. l.25 Let \( a be a real number, and let \( f \) be a function. LATEX realizes that the first \( opens a math environment, so the second \( must be in error. In this case, the line number in the error message is correct. Throughout this book, like nearly everyone else, I use $ to delimit inline math.
Rule
Math environments No blank lines are permitted in a math or displaymath environment.
154
Chapter 7
Typing math
If you violate this rule, LATEX generates an error message, ! Missing $ inserted. $ ... l.7 where the line number points inside the environment. Multiline math environments, such as the examples in Sections 3.4.2–3.4.3, are discussed in Chapter 9.
7.2
Spacing rules In text, the most important spacing rule is that any number of spaces in the source file equals one space in the typeset document. The spacing rule for math mode is even more straightforward.
Rule
Spacing in math LATEX ignores spaces in math. In other words, all spacing in math mode is provided by LATEX. For instance, $a+b=c$ and $a + b = c$ are both typeset as a + b = c. There are two exceptions to this rule: 1. A space indicating the end of a command name is recognized. For instance, in $a \quad b$ LATEX does not ignore the space between \quad and b. 2. If you switch back to text mode inside a math formula with a \text command (see Section 7.4.6), then the text spacing rules apply in the argument of such a command. So, LaTeX provides controls for spaces in typeset math. The spaces you type in math do not affect the typeset document. But keep this tip in mind.
Tip
Format your source file so that it is easy to read.
7.2
155
Spacing rules When typing a source file, the following is good practice: Place \[ and \] on lines by themselves.
Leave spaces before and after binary operations and binary relations, including the equal sign. Indent—by three spaces, for example—the contents of environments so they stand out. Keep a formula on a single line of the source file, if you can. Develop your own style of typing math, and stick with it.
Tip
The spacing after a comma is different in math and text. I use the following rule: If the comma could be replaced with the word “and” or “or”, then I break the formula and move the comma out of the formula. Example 1.
Type “a, b ∈ B” as
$a$, $b \in B$ and not as $a, b \in B$ Example 2.
Type “x = a, b, or c” as
$x = a$, $b$, or $c$ and not as $x = a, b$, or $c$ Compare: x = a, b, or c x = a, b, or c Example 3.
(typed as $x = a$, $b$, or $c$) (typed as $x = a, b$, or $c$) Type “for i = 1, 2, . . . , n” as
“for $i = 1$,~$2$, \dots,~$n$”
Tip
Do not leave a trailing comma in inline math.
156
Chapter 7
Typing math
So do not type If $a = b,$ then but move the comma out.
7.3
Equations An equation is a numbered formula displayed on a single typeset line. Equations are typed in an equation environment. The equation environment and displaymath environment are exactly the same except that the equation environment assigns a number to the displayed formula Z (1)
∞
−∞
e−x dx = 2
√
π
This example is typed as \begin{equation}\label{E:int} \int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi} \end{equation} The \label command in the equation environment is optional. If you use a \label command, the number assigned to the equation can be referenced with the \ref command. So see~(\ref{E:int}) typesets as see (1). Even better, use the \eqref command, which places the parentheses automatically: see~\eqref{E:int} also typesets as see (1). In fact, the \eqref command does more: It typesets the reference upright, even in italicized or slanted text. For more information about crossreferencing, see Section 10.4.2. Analogously, the \upn command forces the use of upright characters for digits, punctuations, parentheses, etc. LATEX numbers equations consecutively. As a rule, equations are numbered consecutively throughout articles, whereas in books, numbering starts from 1 at the start of each chapter. You may also choose to have equations numbered within each section— (1.1), (1.2), . . . , in Section 1; (2.1), (2.2), . . . , in Section 2; and so on—by including the command \numberwithin{equation}{section}
7.4
Basic constructs
157
in the preamble of your document (see Section 10.2). “Manual control” of numbering is discussed in Section 15.5.1, group numbering in Section 9.4.4. The *ed form of the equation environment suppresses numbering, so it is equivalent to the displaymath environment (or the special braces \[ and \]).
Rule
Equation environment No blank lines are permitted within an equation or equation* environment. If you typeset \begin{equation}\label{E:int} \int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi} \end{equation} LATEX generates the familiar, but misleading, error message ! Missing $ inserted.
7.4
Basic constructs A formula is built up by combining various basic constructs. This section discusses the following constructs: Arithmetic operations – Subscripts and superscripts Binomial coefficients Ellipses Integrals Roots Read carefully the basic constructs important for your work. Additional constructs are discussed in subsequent sections.
7.4.1 Arithmetic operations The arithmetic operations are typed pretty much as you would expect. To get a + b, a − b, −a, a/b, and ab, type $a + b$, $a  b$, $a$, $a / b$, $a b$
158
Chapter 7
Typing math
There are two other forms of multiplication and one of division: a · b, a × b, and a ÷ b. They are typed as follows: $a \cdot b$,
$a \times b$,
$a \div b$
In displayed formulas, fractions are usually typed with the \frac command. To get
1 + 2x x + y + xy
type \[
\frac{1 + 2x}{x + y + xy}
\] You can use displaystyle fractions inline with \dfrac, and inlinestyle frac3 + a2 tions in displayed math environments with \tfrac; for example, is typed as 4+b $\dfrac{3 + a^{2}}{4 + b}$ and 3+a2 4+b
is typed as \[ \]
\tfrac{3 + a^{2}}{4 + b}
The \dfrac command is often used in matrices whose entries would look too small with the \frac command. See Formula 20 in the Formula Gallery (Section 7.9) for an example, and Section 8.6.1 for other fraction variants. Subscripts and superscripts Subscripts are typed with _ and superscripts with ^. Remember to enclose the subscripted or superscripted expression in braces: \[
\]
a_{1},\ a_{i_{1}},\ a^{2},\ a^{b^{c}},\ a^{i_{1}},\ a_{i} + 1,\ a_{i + 1},\ a_{1}^{2},\ a^{2}_{1}
typesets as
a1 , ai1 , a2 , ab , ai1 , ai + 1, ai+1 , a21 , a21 c
For ab , type $a^{b^{c}}$, not $a^{b}^{c}$. If you type the latter, you get the error message c
7.4
159
Basic constructs ! Double superscript.
Similarly, abc is typed as $a_{b_{c}}$, not as $a_{b}_{c}$. In many instances, the braces for the subscripts and superscripts could be omitted, but you should type them anyway.
Tip
You may safely omit the braces for a subscript or superscript that is a single digit or letter, as in $a_1$ and $(a + b)^x$, which are typeset as a1 and (a + b)x . Be careful, however. If you have to edit $a_1$ to make it a12 , then the braces can no longer be omitted, you must type $a_{12}$ to obtain a12 because $a_12$ typesets as a1 2. There is one symbol that is automatically superscripted in math mode, the prime, that is, ’. To get f 0 (x), type $f’(x)$. However, to get f 02 you must type $f^{\prime 2}$ 2
Typing ${f’}^{2}$ results in f 0 , with the 2 too high; typing it as $f’^{2}$ causes a double superscript error. Sometimes you may want a symbol to appear superscripted or subscripted by itself, as in the phrase use the symbol
†
to indicate the dualspace
typed as use the symbol ${}^{\dagger}$ to indicate the dualspace where { } is the empty group. The empty group can be used to separate symbols, to terminate commands, or as the base for subscripting and superscripting. The \sb and \sp commands also typeset subscripts and superscripts, respectively, as in $a\sb{1}  a\sp{x + y}$ which produces a1 − ax+y . These commands are seldom used, however, except in the alltt environment (see Section 12.3) and in the Mathematical Reviews of the AMS. For multiline subscripts and superscripts, see Section 7.6.5.
7.4.2 Binomial coefficients Binomials are typeset with the \binom command. Here are two examples shown inline, n2 −1 a 2 b+c and n+1 , and displayed: n2 −1 a 2 and b+c n+1 The latter is typed as
160
Chapter 7 \[
\]
Typing math
\binom{a}{b + c} \text{ and } \binom{\frac{n^{2}  1}{2}}{n + 1}
You can use displaystyle binomials inline with \dbinom, and inlinestyle binoa mials in displayed math environments with \tbinom. For example, is typed b+c as $\dbinom{a}{b + c}$. See Section 8.6.1 for other variants.
7.4.3
Ellipses There are two types of ellipsis in math, the low or ontheline ellipsis, as in F (x1 , x2 , . . . , xn ) and the centered ellipsis, as in x1 + x2 + · · · + xn These two formulas are typed as \[
F(x_{1}, x_{2}, \dots, x_{n})
\] and \[ \]
x_{1} + x_{2} + \dots + x_{n}
LATEX uses the symbol following a \dots command to decide whether to use a low or centered ellipsis. If it fails to make the right decision as in α(x1 + x2 + . . . ) typed as \[
\alpha(x_{1} + x_{2} + \dots)
\] help LATEX by giving the command \ldots for low and \cdots for centered ellipsis. So to get the last formula right, type \[ \]
\alpha(x_{1} + x_{2} + \cdots)
7.4
161
Basic constructs and it typesets correctly: α(x1 + x2 + · · · ) There are five more variants of the \dots command: \dotsc, for an ellipsis followed by a comma \dotsb, for an ellipsis followed by a binary operation or relation \dotsm, for an ellipsis followed by multiplication \dotsi, for an ellipsis with integrals \dotso, for an “other” ellipsis
These commands not only force the ellipsis to be low or centered, but also adjust the spacing. See Section 9.7.1 for an example of vertical dots with the \vdots command and diagonal dots with the \ddots command.
7.4.4 Integrals
R∞ √ 2 You have already seen the formula −∞ e−x dx = π in both inline and displayed forms in the first section of this chapter. The lower limit is typeset as a subscript and the upper limit is typeset as a superscript. To force the limits below and above the integral symbol, use the \limits command. The \nolimits command does the R∞ −x2 √ reverse. To typeset e dx = π, type −∞
$\int\limits_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi}$ See Section 11.5 for a discussion of the intlimits document class option. There are five commands to produce variants of the basic integral symbol: \oint
\iint
\iiint
ZZ
ZZZ
\iiiint
\idotsint
which typeset as I
ZZZZ
Z
Z ···
For complicated bounds, use the \substack command or the subarray environment (see Section 7.6.5).
7.4.5 Roots The \sqrt command produces a square root, for instance,
162
Chapter 7 $\sqrt{5}$ $\sqrt{a + 2b + c^{2}}$
typesets as typesets as
Typing math
√ √5 a + 2b + c2
Here is a more interesting example: v v u s u u r u u t1 + t1 + 1 1 + 1 1 + 1 √1 + · · · 2 3 4 typed as \[ \sqrt{1 + \sqrt{1 + \frac{1}{2}\sqrt{1 + \frac{1}{3} \sqrt{1 + \frac{1}{4}\sqrt{1 + \cdots}}}}} \] For nth roots other √ than the square root, that is, n 6= 2, specify n with an optional argument. To get 3 5, type $\sqrt[3]{5}$. Root refinement √ In g 5, typed as $\sqrt[g]{5}$, the placement of g is not very pleasing. LATEX provides two additional commands to allow you to adjust the position of g: \leftroot moves g left—or right with a negative argument \uproot moves g up—or down with a negative argument You may prefer one of the following variants: √ g 5 typed as $\sqrt[\leftroot{2} \uproot{2} g]{5}$ √ g 5 typed as $\sqrt[\uproot{2} g]{5}$ Experiment with \leftroot and \uproot to find the best spacing. Note that LATEX is very finicky with this optional argument. Typing a space after [, as in $\sqrt[ \uproot{2} g]{5}$, gives the error message ! Package amsmath Error: Invalid use of \uproot. There may also be problems with vertical spacing under the root symbol (see Section 8.4).
7.4.6
Text in math LATEX allows you to include text in formulas with the \text command. The formula A = { x  x ∈ Xi , for some i ∈ I } is typed as
7.4
163
Basic constructs \[
\]
A = \{\, x \mid x \in X_{i}, \text{ for some } i \in I \,\}
Note that you have to leave space before for and after some inside the argument of \text. The argument of the \text command is always typeset in a single line. Sometimes it is more convenient to go into math mode within the argument of a \text command rather than end the \text and start another, as in A = { x  for x large } which is typed as \[ \]
A = \{\, x \mid \text{for $x$ large} \,\}
The \text command correctly sizes its argument to match the context. The formula aleft + 2 = aright is typed as \[ \]
a_{\text{left}} + 2 = a_{\text{right}}
Note that \text typesets its argument in the size and shape of the surrounding text. If you want the text in a formula to be typeset in the document font family (see Section 5.6.2) independent of the surrounding text, use \textnormal{ ... } or {\normalfont ...} For instance, if you have a constant aright , then in a theorem: Theorem 1. The constant aright is recursive in a. The subscript is wrong. To get it right, type the constant as $a_{\normalfont\text{right}}$ Now the theorem typesets as Theorem 1. The constant aright is recursive in a. Any of the text font commands with arguments (see Section 5.6.3) can also be used in math formulas. For instance, \textbf uses the size and shape of the surrounding text to typeset its argument in bold (extended).
164
7.4.7
Chapter 7
Typing math
Building a formula stepbystep It is easy to build up complex formulas from the components described in this section. Try the formula q 3 [ n2 ] i2 X µ(i) 2 (i2 − 1) xi,i+1 i+3 p p 3 ρ(i) − 2 + 3 ρ(i) − 1 3 i=1 Build this formula in several steps. Create a new file in your work folder. Name it formula.tex, type in the following lines, and save it: % File: formula.tex \documentclass{amsart} \usepackage{amssymb,latexsym} \begin{document} \end{document} At present, the file has an empty document environment. Type each part of the formula as an inline or displayed formula within this environment so that you can typeset the document and check for errors. Step 1 We start with n2 . Type the following line into formula.tex: $\left[ \frac{n}{2} \right]$ and test it by typesetting the document. Step 2
Now you can do the sum
[ n2 ] X i=1
For the superscript, you can copy and paste the formula created in Step 1 (without the dollar signs), so that you have \[ \]
\sum_{i = 1}^{ \left[ \frac{n}{2} \right] }
Step 3
Next, do the two formulas in the binomial i+3 i2 xi,i+1 3
Type them as separate formulas in formula.tex: \[ \]
x_{i, i + 1}^{i^{2}}\qquad\left[ \frac{i + 3}{3} \right]
7.4
165
Basic constructs
Step 4 Now it is easy to do the binomial. Piece together the following formula by copying and pasting the previous formulas and dropping the \qquad command: \[ \]
\binom{x_{i,i + 1}^{i^{2}}}{\left[\frac{i + 3}{3}\right]}
which typesets as
xii,i+1 i+3 2
3
Step 5
Next, type the formula under the square root, µ(i) 2 (i2 − 1): 3
$\mu(i)^{ \frac{3}{2} } (i^{2}  1)$ q 3 and then the square root, µ(i) 2 (i2 − 1): $\sqrt{ \mu(i)^{ \frac{3}{2} } (i^{2}  1) }$ Step 6
The two cube roots,
p 3
ρ(i) − 2 and
$\sqrt[3]{ \rho(i)  2 }$ Step 7
p 3 ρ(i) − 1, are easy to type:
$\sqrt[3]{ \rho(i)  1 }$
Now the fraction q
3
µ(i) 2 (i2 − 1) p p 3 ρ(i) − 2 + 3 ρ(i) − 1 is typed, copied, and pasted together as \[ \frac{ \sqrt{ \mu(i)^{ \frac{3}{2}} (i^{2} 1) } } { \sqrt[3]{\rho(i)  2} + \sqrt[3]{\rho(i)  1} } \] Step 8
Finally, the whole formula, [ n2 ] i2 X xi,i+1 i+3 p 3 i=1
3
q
3
µ(i) 2 (i2 − 1) p ρ(i) − 2 + 3 ρ(i) − 1
is formed by copying and pasting the pieces together, leaving only one pair of displayed math delimiters:
166
Chapter 7 \[
\]
Typing math
\sum_{i = 1}^{ \left[ \frac{n}{2} \right] } \binom{ x_{i, i + 1}^{i^{2}} } { \left[ \frac{i + 3}{3} \right] } \frac{ \sqrt{ \mu(i)^{ \frac{3}{2}} (i^{2}  1) } } {\sqrt[3]{\rho(i)2} + \sqrt[3]{\rho(i)  1}}
Note the use of Hierarchical indentation, to keep track of the structure of the formula Spacing to help highlight the braces—some text editors help you balance braces Separate lines for the various pieces of formulas that are more than a line long It is to your advantage to keep your source file readable. LATEX does not care how its input is formatted, and would happily accept the following: \[\sum_{i=1}^{\left[\frac{n}{2}\right]}\binom{x_{i,i+1} ^{i^{2}}}{\left[\frac{i+3}{3}\right]}\frac{\sqrt{\mu(i) ^{\frac{3}{2}}(i^{2}1)}}{\sqrt[3]{\rho(i)2}+\sqrt[3] {\rho(i)1}}\] But this haphazard style not only makes it more difficult for your coauthors or editor to work with your source file, it also makes finding mistakes difficult. Try to find the error in the next version: \[\sum_{i=1}^{\left[\frac{n}{2}\right]} \binom{x_{i,i+1}^{i^{2}}}{\left[\frac{i+3}{3}\right]} \frac{\sqrt{\mu(i)^{\frac{3}{2}}}(i^{2}1)}}{\sqrt[3] {\rho(i)2}+\sqrt[3]{\rho(i)1}}\] Answer: \frac{3}{2 should be followed by }} and not by }}}.
7.5
Delimiters Delimiters are used to enclose some subformulas. In the following formula we use two delimiters: parentheses and square brackets: $[(a*b)+(c*d)]^2$; this typesets as [(a ∗ b) + (c ∗ d)]2 . LATEX knows that parentheses and square brackets are delimiters, and spaces them accordingly. The standard delimiters are shown in Table 7.1. Note that delimiters are math symbols with special spacing rules and you can use them in any way you please, not only in pairs. LATEX does not stop you from typing \uparrow(x)], which typesets as ↑ (x)]. Observe the difference in spacing between a and kak. The first, a, was typed incorrectly as $ a $. As a result, the vertical bars are too far apart. The second
7.5
167
Delimiters Name
Type
left parenthesis right parenthesis left bracket right bracket left brace right brace backslash forward slash left angle bracket right angle bracket vertical line double vertical line left floor right floor left ceiling right ceiling upward double upward downward double downward upanddown double upanddown upperleft corner upperright corner lowerleft corner lowerright corner
( ) [ or \lbrack ] or \rbrack \{ or \lbrace \} or \rbrace \backslash / \langle \rangle  or \vert \ or \Vert \lfloor \rfloor \lceil \rceil \uparrow \Uparrow \downarrow \Downarrow \updownarrow \Updownarrow \ulcorner \urcorner \llcorner \lrcorner
Typeset ( ) [ ] { } \ / h i  k b c d e ↑ ⇑ ↓ ⇓ l m p q x y
Table 7.1: Standard delimiters. was typed correctly using the appropriate delimiter commands: $\ a \$. Here they are again sidebyside, enlarged: a kak.
7.5.1 Stretching delimiters All delimiters, except the four “corners”, can stretch to enclose the subformula. For example, α 1 2 is typed as \[ \]
\left( \frac{1}{2} \right)^{\alpha}
168
Chapter 7
Typing math
The \left and \right commands instruct LATEX to stretch the parentheses. The general construction is \left delim1 and \right delim2 where delim1 and delim2 are chosen from Table 7.1. They are usually, but not always, a matching pair—see the examples below. LATEX inspects the formula between the \left and \right commands and decides what size delimiters to use. The \left and \right commands must be paired in order for LATEX to know the extent of the material to be vertically measured. However, the delimiters need not be the same. If you want to stretch a single delimiter, you have to pair it with a blank delimiter, represented by the \left. and \right. commands. Here are some examples of stretching delimiters: a i a + b
2 b
A , , b , F (x)a 2 , 2 typed as \[
\left \frac{a + b}{2} \right, \quad \left\ A^{2} \right\, \quad \left( \frac{a}{2}, b \right], \quad \left. F(x) \right_{a}^{b}
\] There are also two convenient abbreviations: \left< for \left\langle \right> for \right\rangle The \left and \right commands have one more use. For the delimiters , \, and all the arrows, the same symbol represents the left and right delimiters, which can sometimes cause problems as in Example 2 in Section 8.1.4. In such cases, you should use the \left and \right commands to tell LATEX whether the delimiter is a left or a right delimiter. LATEX also provides the \lvert and \rvert for  as left and right delimiter, and \lVert and \rVert for \.
7.5.2
Delimiters that do not stretch LATEX provides the \big, \Big, \bigg, and \Bigg commands to produce delimiters of larger sizes. These delimiters do not stretch. For example, \[ \]
(\quad \big(\quad \Big(\quad \bigg(\quad \Bigg(
7.5
169
Delimiters typesets as
(
LATEX also provides the more specific \bigl, \Bigl, \biggl, \Biggl, \bigr, \Bigr, \biggr, and \Biggr commands to produce larger left and right delimiters. For integral evaluation, you can choose one of the following: b F (x)
b F (x) a
F (x)ba
a
typed as \[ F(x) ^{b}_{a} F(x) \bigr^{b}_{a} F(x) \Bigr^{b}_{a}
\quad \quad
\]
7.5.3 Limitations of stretching In a number of situations the stretching done by LATEX is not ideal, so you should use a larger sized nonstretching variant. Here are some typical examples: 1. Large operators \[
\]
\left[ \sum_i a_i \right]^{1/p} \quad \biggl[ \sum_i a_i \biggr]^{1/p}
typesets as
"
X i
#1/p ai
X
1/p ai
i
You may prefer the second version with \biggl[ and \biggr]. 2. Groupings \[
\left( (a_1 b_1)  (a_2 b_2) \right) \left( (a_2 b_1) + (a_1 b_2) \right) \quad
170
Chapter 7
Typing math
\bigl( (a_1 b_1)  (a_2 b_2) \bigr) \bigl( (a_2 b_1) + (a_1 b_2) \bigr) \] typesets as ((a1 b1 ) − (a2 b2 )) ((a2 b1 ) + (a1 b2 ))
(a1 b1 ) − (a2 b2 ) (a2 b1 ) + (a1 b2 )
You may prefer the clearer groupings provided by \bigl( and \bigr). 3. Inline formulas 0 The delimiters produced by \left and \right use too much interline space in db 0 , typed as \left\lvert \frac{b’}{d’} \right\rvert Use b0 \bigl and \bigr to produce delimiters that fit within the normal line spacing: 0 , typed as d
\bigl\lvert \frac{b’}{d’} \bigr\rvert
7.5.4
Delimiters as binary relations The symbol  can be used as a delimiter, as in x + y, and also as a binary relation, as in { x ∈ R  x2 ≤ 2 }. As a binary relation it is typed as \mid. The previous formula is typed as $\{\, x \mid x^{2} \leq 2 \,\}$ \bigm and \biggm produce larger variants, with spacing on either side like binary relations. For example, Z x 2 x t dt ≤ 5 0
is typed as \[ \left\{\, x \biggm\int_{0}^x t^{2}\, dt\leq 5\,\right\} \]
7.6
Operators You cannot just type sin x to typeset the sine function in math mode. Indeed, $sin x$
7.6
171
Operators produces sinx instead of sin x, as you intended. Type this function as $\sin x$
The \sin command prints sin with the proper style and spacing. LATEX calls \sin an operator or loglike function.
7.6.1 Operator tables There are two types of operators: 1. Operators without limits, such as \sin 2. Operators with limits, such as \lim, that take a subscript in inline mode and a “limit” in displayed math mode. For example, limx→0 f (x) = 1 is typed as $\lim_{x \to 0}
f(x) = 1$
The same formula displayed,
lim f (x) = 1
x→0
is typed as \[
\lim_{x \to 0} f(x) = 1
\] The operators are listed in Tables 7.2 and 7.3 (see also Section B.6). The entries in the last two rows of Table 7.3 can be illustrated by lim
x→0
lim
x→0
lim −→
x→0
lim ←−
x→0
which are typed as \[
\]
\varliminf_{x \to 0} \quad \varinjlim_{x \to 0} \quad
\varlimsup_{x \to 0} \quad \varprojlim_{x \to 0}
The following examples illustrate some more entries from Table 7.3: inj lim x→0
lim inf x→0
lim sup x→0
proj lim x→0
These operators were typed as \[ \injlim_{x \to 0} \quad \liminf_{x \to 0} \quad \limsup_{x \to 0} \quad \projlim_{x \to 0} \]
172
Chapter 7
Typing math
You can force the limits in a displayed formula into the subscript position with the \nolimits command. For example, the formulas inj limx→0
lim inf x→0
lim supx→0
proj limx→0
are typed as \[
\]
\injlim\nolimits_{x \to 0} \quad \liminf\nolimits_{x \to 0} \quad \limsup\nolimits_{x \to 0} \quad \projlim\nolimits_{x \to 0}
Type
Typeset Type
Typeset
Type
Typeset
Type
Typeset
\arccos \arcsin \arctan \arg \cos \cosh
arccos arcsin arctan arg cos cosh
cot coth csc deg dim exp
\hom \ker \lg \ln \log \sec
hom ker lg ln log sec
\sin \sinh \tan \tanh
sin sinh tan tanh
\cot \coth \csc \deg \dim \exp
Table 7.2: Operators without limits.
Type
Typeset
Type
Typeset
\det \gcd \inf \lim \liminf \injlim
det gcd inf lim lim inf inj lim
\limsup \max \min \Pr \sup \projlim
lim sup max min Pr sup proj lim
\varliminf
lim
\varlimsup
lim
\varinjlim
lim −→
\varprojlim
lim ←−
Table 7.3: Operators with limits.
7.6
173
Operators
7.6.2 Defining operators The powerful \DeclareMathOperator command defines a new operator: \DeclareMathOperator{\opCommand }{opName } Invoke the new operator with \opCommand, which is then typeset with opName. The \DeclareMathOperator command must be placed in the preamble. For example, to define the operator Truncat, invoked by the command \Trunc, place this in the preamble: \DeclareMathOperator{\Trunc}{Truncat} An operator is typeset in math roman with a little space after it, so $\Trunc A$ typesets as Truncat A. The second argument is typeset in math mode but  and * are typeset as they would be in text. Here are some more examples. Define in the preamble two operators: \DeclareMathOperator{\Trone}{Trunc_{1}} \DeclareMathOperator{\Ststar}{Starone*} Then in the body of the article $\Trone A$ is typeset as Truncat1 A $\Ststar A$ is typeset as Starone* A To define an operator with limits, use the *ed form \DeclareMathOperator*{\doublesum}{\sum\sum} and then (see Section 7.6.5 for multiline subscripts) \[
\]
\doublesum_{\begin{subarray}{l} i^2+j^2 = 50\\ i,\ j \leq 10 \end{subarray}} \frac{x^i + y^j}{(i + j)!}
typesets as
X X xi + y j (i + j)! 2 2
i +j =50 i, j≤10
7.6.3 Congruences \mod is a special operator used for congruences. Congruences are usually typeset using the \pmod or \pod variant. There is also the \bmod command, which is used as a binary operation. All four commands are shown in Figure 7.4. See Sections 15.1.2 and 15.1.8 for a discussion of related userdefined commands.
174
Chapter 7 Type
Typeset
$a $a $a $a
a ≡ v mod θ a mod b a ≡ v (mod θ) a ≡ v (θ)
\equiv v \mod{\theta}$ \bmod b$ \equiv v \pmod{\theta}$ \equiv v \pod{\theta}$
Typing math
Table 7.4: Congruences.
7.6.4
Large operators Here is a sum typeset inline,
Pn
i=1
x2i , and displayed, n X
x2i
i=1
In the latter form, the sum symbol is larger. Operators that behave in this way are called large operators. Table 7.5 gives a complete list of large operators. You can use the \nolimits command if you wish to show the limits of large operators as subscripts and superscripts in a displayed math environment. The formula G X=a m
is typed as \[ \]
\bigsqcup\nolimits_{ \mathfrak{m} } X = a
You can use the \limits command if you wish to show the limits of large operators below and above the operator symbol in an inline math environment. For example, F X = a is typed as m
$\bigsqcup\limits_{ \mathfrak{m} } X = a$ Sums and products are very important constructs. The examples d
X z d − z0d = z0k−1 z d−k z − z0 k=1
are typed as \[
\frac{z^{d}  z_{0}^{d}} {z  z_{0}} =
and
(T n )0 (x0 ) =
n−1 Y k=0
T 0 (xk )
7.6
175
Operators Type
Inline Rb
\int_{a}^{b}
a
a
\prod_{i=1}^{n} \coprod_{i=1}^{n} \bigcap_{i=1}^{n} \bigcup_{i=1}^{n} \bigwedge_{i=1}^{n} \bigvee_{i=1}^{n} \bigsqcup_{i=1}^{n} \biguplus_{i=1}^{n} \bigotimes_{i=1}^{n} \bigoplus_{i=1}^{n} \bigodot_{i=1}^{n} \sum_{i=1}^{n}
a
a
a
RRRR b a
R
···
Rb a
b
a
ZZZ
RRR b
\iiint_{a}^{b}
b
ZZ
RR b
\iint_{a}^{b}
\idotsint_{a}^{b}
a
I
Hb
\oint_{a}^{b}
\iiiint_{a}^{b}
Displayed Z b
a
ZZZZ
Z ··· n Y
`n
i=1 n a
Tn
i=1 n \
Sn
i=1 n [
Vn
i=1 n ^
Wn
i=1 n _
Fn
i=1 n G
Un
i=1 n ]
Nn
n O
Ln
n M
Jn
n K
Pn
n X
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
Table 7.5: Large operators.
b
a
Z
Qn
i=1
b
i=1
i=1
i=1
i=1
i=1
a
b
176
Chapter 7
\]
7.6.5
Typing math
\sum_{k = 1}^{d} z_{0}^{k  1} z^{d  k} \text{\quad and\quad} (T^{n})’(x_{0}) = \prod_{k=0}^{n  1} T’(x_{k})
Multiline subscripts and superscripts The \substack command provides multiline limits for large operators. For instance, X x2i i (x + 1)^{2}  \alpha \,\}
\] The spacing of the symbols in the formula varies. In xβ, the two symbols are very close. In x ∈ X, there is some space around the ∈, and in x + 1, there is somewhat less space around the +. There is a little space after { and before }—not quite enough for this formula, which is why the thinspace commands (\,) were added.
8.1.1
Classification LATEX classifies symbols into several categories or types and spaces them accordingly. In the formula A = { x ∈ X  xβ ≥ xy > (x + 1)2 − α } we find Ordinary math symbols: A, x, X, β, and so on Binary relations: =, ∈, , ≥, and > Binary operations: + and − Delimiters: {, }, (, and ) As a rule, you do not have to be concerned with whether or not a given symbol in a formula, say ×, is a binary operation. LATEX spaces the typeset symbol correctly.
8.1.2
Three exceptions There are three symbols with more than one classification: +, −, and . + or − could be either a binary operation, for instance, a − b, or a sign, for instance, −b.
Rule
+ and − + or − are binary operations when preceded and followed by a symbol or an empty group ({ }).
8.1
189
Spacing of symbols So, for instance, in (A + BC)x +
Cy = 0,
Ex + (F + G)y = 23. which is typed as (see the alignat* environment in Section 9.5.4) \begin{alignat*}{2} (A + B C)x &{}+{} &C &y = 0,\\ Ex &{}+{} &(F + G)&y = 23. \end{alignat*} we use the empty groups, { }, to tell LATEX that the second + in line 1 and the first + in line 2 of the formula are binary operations. If we leave out the empty groups, and type instead \begin{alignat*}{2} (A + B C)x &+ &C &y = 0,\\ Ex &+ &(F + G)&y = 23. \end{alignat*} we get (A + BC)x+
Cy = 0,
Ex+(F + G)y = 23. Another illustration is given later in this section using the \phantom command. This problem often arises in split formulas, for example if the formula is split just before a + or −, you should start the next line with {}+ or {}. See Section 9.3 for examples. The  symbol can play several different roles in a math formula, so LATEX provides separate commands to specify the symbol’s meaning.
Rule
The four roles of the  symbol 
ordinary math symbol
\mid
binary relation
\left
left delimiter
\right right delimiter Note the differences between the spacing in ab, typed as $a  b$, and in a  b, typed as $a \mid b$.
190
Chapter 8 Name
Width
1 mu (math unit) thinspace medspace thickspace interword space 1 em 2 em
Short \, \: \; \
More math
Long \mspace{1mu} \thinspace \medspace \thickspace \quad \qquad
Negative space 1 mu thinspace medspace thickspace
\!
\mspace{1mu} \negthinspace \negmedspace \negthickspace
Table 8.1: Math spacing commands.
8.1.3
Spacing commands There are some situations where LATEX cannot typeset a formula properly and you have to add spacing commands. Luckily, LATEX provides a variety of spacing commands, listed in Table 8.1. The \neg commands remove space by “reversing the print head”. The \quad and \qquad commands are often used to adjust aligned formulas (see Chapter 9) or to add space before text in a math formula. The size of \quad (= 1 em) and \qquad (= 2 em) depends on the current font. The \, and \! commands are the most useful for fine tuning math formulas, see some examples in the Formula Gallery and in the next section. The \mspace command and the math unit mu provides you with even finer control. 18 mu = 1 em, defined in Section 5.8.3. For example, \mspace{3mu} adds a space that is 1/6 em long. There is an interesting use of mu on page 109.
8.1.4
Examples The Formula Gallery in Section 7.9 starts out with an example that shows the importance of fine tuning. In set notation, when using \mid for “such that”, thin spaces are inserted just inside the braces. Some more examples of fine tuning follow. One more example can be found in Section 8.2.1. Rπ Example 1 In Section 3.3, we type the formula 0 sin x dx = 2 as $\int_{0}^{\pi} \sin x \, dx = 2$ Notice the thinspace spacing command \, between R π \sin x and dx. Without the command, LATEX would have crowded sin x and dx: 0 sin xdx = 2.
8.1
191
Spacing of symbols
Example 2  − f (x), typed as $f(x)$, is spaced incorrectly. − becomes a binary operation by the + and − rule. To get the correct spacing, as in −f (x), type $\leftf(x)\right$. This form tells LATEX that the first  is a left delimiter, by the  rule, and therefore − is the unary minus sign, not the binary subtraction operation. √ Example 3 In 5side, typed as $\sqrt{5} \text{side}$ √ 5 is too close to side. So type it as $\sqrt{5} \, \text{side}$ √ which typesets as 5 side. Example 4 In sin x/ log n, the division symbol / is too far from log n, so type $\sin x / \! \log n$ which prints sin x/log n. √ Example 5 In f (1/ n), typed as $f(1 / \sqrt{n})$ the square root almost touches the closing parenthesis. To correct it, type $f(1 / \sqrt{n}\,)$ √ which typesets as f (1/ n ). There is one more symbol with special spacing: the \colon command, used for formulas such as f : A → B (typed as $f \colon A \to B$). Observe that $f: A \to B$ typesets as f : A → B. The spacing is awful. See Section 8.2.3 on how to declare the type of a symbol.
8.1.5 The phantom command The \phantom{argument } command (introduced for text in Section 5.8.1) produces a space in a formula equivalent to the space that would be occupied by its typeset argument. This command is one of the most powerful tools available to us for fine tuning alignments. Here are two simple illustrations:
1 A= 2 −2
3 1 1 1 2 −1
192
Chapter 8
More math
typed as \[
A = \begin{pmatrix} \phantom{}1 & \phantom{}3 & \phantom{}1\\ \phantom{}2 & \phantom{}1 & \phantom{}1\\ 2 & \phantom{}2 & 1\\ \end{pmatrix}
\] and a+b+c+d
= 0,
c + d + e = 5. typed as \begin{align*} a + b + c & + d \phantom{ {}+e } = 0,\\ c & + d + e = 5. \end{align*} Note that \phantom{+e} yields incorrect spacing by the + and − rule: a+b+c+d
= 0,
c + d + e = 5. See Section 9.6.2 for an additional example.
8.2
Building new symbols No matter how many math symbols LATEX provides—see a complete listing in Appendix B—users always seem to want more.1 LATEX gives you excellent tools to build new symbols from existing ones.
8.2.1
Stacking symbols u
To place any symbol above, or below, any other, for instance, ∼, use the \overset command. It takes two arguments—the first argument is set in a smaller size above the second argument. The spacing rules of the symbol in the second argument remain valid, u i.e., the type remains the same. Since ∼ is a binary relation, so is ∼. The \underset command is the same except that the first argument is set under the second argument. For example, α α α a X ai ai ·
are typed as 1 The
best is yet to come, see Section H.3.
8.2
193
Building new symbols \[
\overset{\alpha}{a} \underset{\boldsymbol{\cdot}}{X} \overset{\alpha}{ a_{i} } \overset{\alpha}{a}_{i}
\qquad \qquad \qquad
\] α
For the \boldsymbol command, see Section 8.3.3. Note that in the third example, ai , the α seems to be sitting too far to the right but the fourth example corrects that. You can also use these commands with binary relations, as in def
f (x) = x2 − 1 which is typed as \[ \]
f(x) \overset{ \text{def} }{=} x^{2}  1 def
Since = is a binary relation, = becomes a binary relation, as shown by the spacing on either side. Here is another example, a u c l e + + b d f which is typed as \[
\]
\frac{a}{b} \overset{u}{+} \frac{c}{d} \overset{l}{+} \frac{e}{f} u
l
Note that + and + are properly spaced as binary operations. As we discuss in Section 7.4.6, the safer definitions for these examples are \[ \]
f(x) \overset{ \normalfont\text{def} }{=} x^{2}  1
and \[
\]
\frac{a}{b} \overset{\normalfont u}{+} \frac{c}{d} \overset{\normalfont l}{+} \frac{e}{f}
194
8.2.2
Chapter 8
More math
Negating and sidesetting symbols You can negate with the \not command; for instance, a 6∈ b and a 6= b are typed as $a \not\in b$ and $a \not= b$, respectively. It is preferable, however, to use the negated symbols ∈, / typed as $\notin$, and 6=, typed as $\ne$. See the negated binary relations table in Section B.2. For instance, “a does not divide b”, a  b, should be typed as $a \nmid b$, not as $a \not\mid b$, which typesets as a 6 b. In Section 8.2.3, you learn how to improve a 6 b to a 6  b, typed as $a \mathrel{\not} b$. However, $a \nmid b$ is still best. LATEX provides the \sideset command to set symbols at the corners of large operators other than the “corners” (the last four delimiters in Table 7.1). This command takes three arguments: \sideset{ _{ll }^{ul } }{ _{lr }^{ur } }{large_op } where ll stands for the symbol to be placed at the lower left, ul for upper left, lr for lower right, and ur for upper right; large op is a large operator. These two examples, Yc eY and a
are typed as \[
\sideset{}{_{a}^{c}}{\prod}\text{ and } \sideset{^{e}}{}{\prod}
\] Note that the two first arguments are compulsory, although one or the other may be empty, while the third argument must contain the large operator. Here is a more meaningful example: \[ \]
\sideset{}{’}{\sum}_{\substack{ i < 10\\ j < 10 } } x_{i}z_{j}
is typeset as
X0
xi zj
i
)
−
α
}
196
8.3.1
Chapter 8
More math
Math alphabets The letters and digits typed in a math formula come from math alphabets. LATEX’s default math alphabet—the one you get if you do not ask for something else—is Computer Modern math italic for letters. In the formula x2 ∨ y3 = α, the characters x and y come from this math alphabet. The default math alphabet for digits is Computer Modern roman and the digits 2 and 3 in this formula are typeset in Computer Modern roman. LATEX has a number of commands to switch type style in math. The two most important commands select the bold and italic versions: Command
Math alphabet
Produces
\mathbf{a} \mathit{a}
math bold math italic
2 Greek gammas, γ and Γ 2 Greek gammas, γ and Γ
These commands change the style of letters, numbers, and upper case Greek characters. But beware of the pitfalls. For instance, in \mathit{leftside} the hyphen typesets as a minus: left − side. There are four more commands that switch math alphabets: Command
Math Alphabet
Produces
\mathsf{a} \mathrm{a} \mathtt{a} \mathnormal{a}
math sans serif math roman math typewriter math italic
2 Greek gammas, γ and Γ 2 Greek gammas, γ and Γ 2 Greek gammas, γ and Γ Greek gammas, γ and Γ
Math roman is used in formulas for operator names, such as sin in sin x, and for text. For operator names, you should use the \DeclareMathOperator command or the *ed version, which sets the name of the operator in math roman, and also provides the proper spacing (see Section 8.2.3). For text, you should use the \text command (see Section 7.4.6). The \mathnormal command switches to the default math alphabet, but this command is seldom used in practice. The Computer Modern fonts include a math bold italic alphabet, but LATEX does not provide a command to access it.
Rule
Math alphabets vs. text alphabets Do not use text alphabets in a math formula, except in the argument of a \text command.
8.3
Math alphabets and symbols
197
It may not be easy for you to see the difference, but some things will not look right or may not align properly.
8.3.2 Math symbol alphabets You may have noticed that α was not classified as belonging to an alphabet in the example at the beginning of this section. Indeed, α is treated by LATEX as a math symbol rather than as a member of a math alphabet. You cannot italicize or slant it, nor is there a sans serif version. There is a bold version, but you must use the \boldsymbol command to produce it. For instance, αβ , is typed as $\boldsymbol{\alpha}_{\boldsymbol{\beta}}$ Note that β appears in a small size in αβ . Four “alphabets of symbols” are built into LATEX. Greek The examples α, β, and Γ are typed as $\alpha$, $\beta$, $\Gamma$ See Section B.1 for the symbol table. Calligraphic an uppercaseonly alphabet invoked with the \mathcal command. The examples A, C, and E are typed as $\mathcal{A}$, $\mathcal{C}$, $\mathcal{E}$ Euler Fraktur invoked by the \mathfrak command. The examples n, p, N, and P are typed as $\mathfrak{n}$, $\mathfrak{p}$, $\mathfrak{N}$, $\mathfrak{P}$ Blackboard bold uppercaseonly math alphabet, invoked with \mathbb. The examples A, B, C are typed as $\mathbb{A}$, $\mathbb{B}$, $\mathbb{C}$
8.3.3 Bold math symbols In math, most characteristics of a font are specified by LATEX. One exception is boldface. To make a letter from a math alphabet within a formula bold, use the \mathbf command. For instance, in we choose the vector v
198
Chapter 8
More math
v is produced by $\mathbf{v}$. To obtain bold math symbols, use the \boldsymbol command. For example, the bold symbols 5 α Λ A → A are typed as \[
\boldsymbol{5} \quad \boldsymbol{\alpha} \quad \boldsymbol{\Lambda}\quad\boldsymbol{\mathcal{A}} \quad \boldsymbol{\to} \quad \boldsymbol{A}
\] Note that \boldsymbol{A} typesets as A, a bold math italic A. To get an upright A, type $\mathbf{A}$. 5 did not really need \boldsymbol; \mathbf{5} gives the same result. To make an entire formula bold, use the \mathversion{bold} command, as in {\mathversion{bold} $a \equiv c \pod{\theta}$} which typesets as a ≡ c (θ). Note that the \mathversion{bold} command is given before the formula. To typeset AMS, type $\boldsymbol{ \mathcal{A} } \boldsymbol{ \mathcal{M} } \boldsymbol{ \mathcal{S} }$ or $\boldsymbol{ \mathcal{AMS} }$ or {\mathversion{bold} $\mathcal{AMS}$} Within the scope of \mathversion{bold}, you can undo its effect with \mathversion{normal} Not all symbols have bold variants. For example, if you type $\sum \quad \boldsymbol{\sum}$ P P you get , two identical symbols. If you want to obtain a bold version, use the poor man’s bold invoked by the \pmb command. This command typesets the symbol three times very close to one another producing a bold symbol of some quality. Note that \pmb does destroy the type of the symbol, \pmb{\sum} is no longer spaced like a large operator. To make it into a large operator, declare in the preamble
8.3
199
Math alphabets and symbols \DeclareMathOperator{\boldsum}{\pmb{\sum}} and \DeclareMathOperator*{\boldsumlim}{\pmb{\sum}} Compare the following four variants of sum: n X
i2
X
n 2 i=1 i
Xn
i=1
i=1
i2
n X
i2
i=1
The first sum is typed (in displayed math mode) as \sum_{i = 1}^{n} i^{2} The second uses poor man’s bold, but does not declare the result to be a large operator: \pmb{\sum}_{i = 1}^{n} i^{2} The third uses the math operator declared: \boldsum_{i = 1}^{n} i^{2} The fourth uses the math operator with limit declared: \boldsumlim_{i = 1}^{n} i^{2}
8.3.4 Size changes There are four math font sizes, invoked by the command declarations \displaystyle, normal size for displayed formulas \textstyle, normal size for inline formulas \scriptstyle, normal size for subscripted and superscripted symbols \scriptscriptstyle, normal size for doubly subscripted and superscripted symbols These commands control a number of style parameters in addition to the size. Compare the two fractions 1 1 1 2+ 1 3 2+ 3 which are typed as \[ \frac{1}{\displaystyle 2 + \frac{1}{3}} \quad \frac{1}{ 2 + \frac{1}{3} } \]
200
8.3.5
Chapter 8
More math
Continued fractions In addition to the \frac, \dfrac, and \tfrac commands (see Section 7.4.1), LATEX makes typesetting continued fractions even easier by providing the \cfrac command. The \cfrac command takes an optional argument, l or r, to place the numerator on the left or on the right. For example,
2+
1
1
1 3 + ···
2+
1 3 + ···
is typed as \[ \cfrac{1}{ 2 + \cfrac{1}{3 + \cdots} } \qquad \cfrac[l]{1}{2 + \cfrac[l]{1}{3 + \cdots}} \]
8.4
Vertical spacing As a rule, all horizontal and vertical spacing in a math formula is done by LATEX. Nevertheless, you often need to adjust horizontal spacing (see Section 8.1). There is seldom a need to adjust vertical spacing, but there are a few exceptions. √ √ The formula a + b does not look quite right, because the square roots are not uniform. You can correct this with \mathstrut commands, which inserts an invisible vertical space: $\sqrt{\mathstrut a} + \sqrt{\mathstrut b}$ p p typesets as a + b. See Section 5.9.5 for struts in general. Another way to handle this situation is with the \vphantom (vertical phantom) command, which measures the height of its argument and places a math strut of that height into the formula. So $\sqrt{\vphantom{b} a} + \sqrt{b}$ √ √ also prints uniform square roots, a + b. The \vphantom method is more versatile than the previous one. Here is a more complicated example from a recent research article: [ [ Θi = Θ( a ∧ b, a ∧ b)  a, b ∈ Bi ∨ Θ( a ∨ b, a ∨ b )  a, b ∈ Bi , typed as \[ \Theta_i = \bigcup \big(\, \Theta (\,\overline{a
8.5
201
Tagging and grouping \wedge b},\overline{\vphantom{b}a} \wedge \overline{b}) \mid a,\ b \in B_i \,\big) \vee \bigcup \big(\, \Theta(\,\overline{a \vee b}, \overline{\vphantom{b}a} \vee \overline{b} \,) \mid a,\ b \in B_i \,\big), \]
Another useful command for vertical spacing is the \smash command. It directs LATEX to pretend that its argument does not protrude above or below the line in which it is typeset. For instance, the two lines of this admonition: It is very important that you memorize the integral R which will appear on the next test.
1 f (x) dx
= 2g(x) + C,
are too far apart because LATEX had to make room for the fraction. However, in this instance, the extra vertical space is not necessary because the second line is very short. To correct this, place the formula in the argument of a \smash command: It is \emph{very important} that you memorize the integral $\smash{\frac{1}{\int f(x) \, dx}} = 2g(x) + C$, which will appear on the next test. LATEX produces the following: It is very important that you memorize the integral R which will appear on the next test.
1 f (x) dx
= 2g(x) + C,
An optional argument to the \smash command controls which part of the formula is ignored, t to smash the top, and b to smash the bottom.
8.5
Tagging and grouping You can attach a name to an equation using the \tag command. In the equation or equation* environments, \tag{name } attaches the tag name to the equation—name is typeset as text. The tag replaces the number. Recall that the numbering of an equation is relative, that is, the number assigned to an equation is relative to the placement of the equation with respect to other equations in the document. An equation tag, on the other hand, is absolute—the tag remains the same even if the equation is moved.
202
Chapter 8
More math
If you want to reference the number generated by LATEX for an equation, then you have to attach a \label{name } command. You reference the number with the \ref{name } or \eqref{name } command. Note that an equation may contain both a tag and a label. The tag is typeset and the label can be used for page referencing with the \pageref command (see Section 10.4.2). If there is a tag, the equation and the equation* environments are equivalent. For example, Z ∞ √ 2 (Int) e−x dx = π −∞
may be typed as \begin{equation*} \int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi}\tag{Int} \end{equation*} or \begin{equation} \int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi}\tag{Int} \end{equation} or \[
\int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi}\tag{Int}
\] Note that \label works in a starred display math environment if a tag is present. The \tag* command is the same as \tag except that it does not automatically enclose the tag in parentheses. To get Z ∞ √ 2 A–B e−x dx = π −∞
type \begin{equation} \int_{\infty}^{\infty} e^{x^{2}} \, dx = \sqrt{\pi} \tag*{AB} \end{equation}
8.5
203
Tagging and grouping Tagging allows numbered variants of equations. For instance, the equation (1)
A[2] B [2] ∼ = (A B)[2]
may need a variant: (10 )
Ah2i B h2i ≡ (A B)h2i
If the label of the first equation is E:first, then the second equation may be typed as follows: \begin{equation}\tag{\ref{E:first}$’$} A^{\langle 2 \rangle} \diamond B^{\langle 2\rangle} \equiv (A \diamond B)^{\langle 2 \rangle} \end{equation} Such a tag is absolute in the sense that it does not change if the equation is moved. But if it references a label and the number generated by LATEX for the label changes, the tag changes. In contrast, grouping applies to a group of adjacent equations. Suppose the last equation was numbered (1) and the next group of equations is to be referred to as (2), with individual equations numbered as (2a), (2b), and so on. Enclosing these equations in a subequations environment accomplishes this goal. For instance, (1a)
A[2] B [2] ∼ = (A B)[2]
and its variant (1b)
Ah2i B h2i ≡ (A B)h2i
are typed as \begin{subequations}\label{E:joint} \begin{equation}\label{E:original} A^{[2]} \diamond B^{[2]} \cong (A \diamond B)^{[2]} \end{equation} \begin{equation}\label{E:modified} A^{\langle 2 \rangle} \diamond B^{\langle 2\rangle} \equiv (A \diamond B)^{\langle 2\rangle} \end{equation} \end{subequations} Referring to these equations, you find that \eqref{E:joint} resolves to (1) \eqref{E:original} resolves to (1a)
204
Chapter 8
More math
\eqref{E:modified} resolves to (1b) Note that in this example, references to the second and third labels produce numbers, (1a) and (1b), that also appear in the typeset version. The group label, E:joint, references the entire group, but (1) does not appear in the typeset version unless referenced. A subequations environment can contain the multiline math constructs discussed in Chapter 9 (see Section 9.4.4).
8.6
Miscellaneous
8.6.1
Generalized fractions The generalized fraction command provides the facility to typeset many variants of a+b fractions and binomials, such as c and a+b c . The syntax is \genfrac{leftdelim }{rightdelim }{thickness }{mathstyle } {numerator }{denominator } where leftdelim is the left delimiter for the formula (default: none) rightdelim is the right delimiter for the formula (default: none) thickness is the thickness of the fraction line, in the form x pt (default: the normal weight, 0.4pt), for instance, 12pt for 12 point width mathstyle is one of – 0 for \displaystyle – 1 for \textstyle – 2 for \scriptstyle – 3 for \scriptscriptstyle – Default: Depends on the context. If the formula is being set in display style, then the default is 0, and so on numerator is the numerator denominator is the denominator All arguments must be specified. The empty argument, {}, gives the default value. Examples 1. \frac{numerator }{denominator } is the same as
8.6
205
Miscellaneous \genfrac{}{}{}{}{numerator }{denominator } 2. \dfrac{numerator }{denominator } is the same as \genfrac{}{}{}{0}{numerator }{denominator } 3. \tfrac{numerator }{denominator } is the same as \genfrac{}{}{}{1}{numerator }{denominator } 4. \binom{numerator }{denominator } is the same as \genfrac{(}{)}{0pt}{}{numerator }{denominator } 5. Here are some more examples: a+b c
a+b c
a+b
a+b
c
c
a+b c
a+b c
typed as \[
\frac{a + b}{c} \quad \genfrac{}{}{1pt}{}{a + \genfrac{}{}{1.5pt}{}{a \genfrac{}{}{2pt}{}{a + \genfrac{[}{]}{0pt}{}{a \genfrac{]}{[}{0pt}{}{a
b}{c} + b}{c} b}{c} + b}{c} + b}{c}
\quad \quad \quad \quad
\] You can choose the delimiters from Table 7.1. If a \genfrac construct is used repeatedly, you should name it. See Section 15.1 for userdefined commands.
8.6.2 Boxed formulas The \boxed command puts its argument in a box, as in Z (2) typed as
∞
−∞
e−x dx = 2
√
π
206
Chapter 8
More math
\begin{equation} \boxed{ \int_{\infty}^{\infty} e^{x^{2}}\, dx = \sqrt{\pi} } \end{equation} The \boxed command can also be used in the argument of a \text command. Note that \fbox{Hello world} and $\boxed{\text{Hello world}}$ produce the same Hello world . Morten Høgholm’s mathtools package (see Section E.1 on how to get it), contains many variants of boxes.
CHAPTER
9 Multiline math displays
9.1
Visual Guide LATEX is about typesetting math. It knows a lot about typesetting inline formulas, but not much about how to display a multiline formula to best reflect its meaning in a visually pleasing way. So you have to decide the visual structure of a multiline formula and then use the tools provided by LATEX to code and typeset it. For many mathematical documents the three constructs of Chapter 3 suffice: simple and annotated alignments, and the cases construct. To help you choose the appropriate tool for more complicated constructs, we start by introducing the basic concepts and constructions with the Visual Guide shown on the next page.
208
Chapter 9
Multiline math displays
Adjusted environments (x1 x2 x3 x4 x5 x6 )2
x1 x2 + x21 x22 + x3
+ (x1 x2 x3 x4 x5 + x1 x3 x4 x5 x6 + x1 x2 x4 x5 x6 + x1 x2 x3 x5 x6 )2
x1 x3 + x21 x23 + x2 x1 x2 x3
+ (x1 x2 x3 x4 + x1 x2 x3 x5 + x1 x2 x4 x5 + x1 x3 x4 x5 )2
gather one column, centered
multline flush left, centered, flush right
Adjusted subsidiary environments
1 0 ... 0 1 . . . .. .. . . . . . 0 0 ...
0 0 .. . 1
matrix multicolumn, centered
a+b+c uv a+b u+v
x−y z
27 134
2 if x < 0; −x , f (x) = α + x, if 0 ≤ x ≤ 1; 2 x , otherwise.
cases array columns flush left multicolumn each column adjusted independently
Aligned environments f (x) = x + yz h(x) = xy + xz + yz
g(x) = x + y + z k(x) = (x + y)(x + z)(y + z)
f (x) = x + yz h(x) = xy + xz + yz
align multicolumn, aligned
g(x) = x + y + z k(x) = (x + y)(x + z)(y + z)
flalign multicolumn, aligned
(17)
a11 x1 + a12 x2 + a13 x3
(18)
a21 x1 + a22 x2
(19)
a31 x1
= y1 + a24 x4 = y2
+ a33 x3 + a34 x4 = y3
alignat multicolumn, aligned
Aligned subsidiary environment i
(3.4)
j
0 = h. . . , 0, . . . , d, . . . , 0, . . . i ∧ h. . . , 0, . . . , a, . . . , 0, . . . i j
≡ h. . . , 0, . . . , a, . . . , 0, . . . i (mod Θ)
split one column, aligned
Figure 9.1: The Visual Guide for multiline math formulas.
9.1
209
Visual Guide
9.1.1 Columns Multiline math formulas are displayed in columns. The columns are either adjusted, that is, centered, or set flush left or right, or aligned, that is, an alignment point is designated for each column and for each line. Moreover, the columns are either separated by the intercolumn space or adjacent with no separation. One column As in Chapter 3, we start with a simple align: r2 = s2 + t2 , 2u + 1 = v + wα This is a single column, aligned at the = signs, and coded with the align environment (see Section 3.4.2). Two columns The annotated align, coded with the align environment, x = x ∧ (y ∨ z)
(by distributivity)
= (x ∧ y) ∨ (x ∧ z)
(by condition (M))
=y∨z has two columns. The first column is aligned like our example of simple align, but the second column is aligned flush left. There is a sizeable intercolumn space.
9.1.2 Subsidiary math environments The cases example in Chapter 3:
f (x) =
2 −x ,
α + x, x 2 ,
if x < 0; if 0 ≤ x ≤ 1; otherwise.
introduces a new concept. The part of the formula to the right of = is a multiline construct. This is an example of a subsidiary math environment that can only be used inside another math environment. It creates a “large math symbol”, in this case 2 if x < 0; −x , α + x, x 2 ,
if 0 ≤ x ≤ 1; otherwise.
210
Chapter 9
Multiline math displays
So the cases example: f (x) = large math symbol is a single line displayed formula, where “large math symbol” is replaced by the cases construct.
9.1.3
Adjusted columns An adjusted column is either set centered, or flush left, or flush right. This may happen by default, built into the environment, or so specified in the code. For instance, in the displayed formula x1 x2 + x21 x22 + x3 , x1 x3 + x21 x23 + x2 typeset with the gather environment, by default all the lines are centered. On the other hand, in 1 100 115 201 0 1 coded with the array subsidiary math environment, the first column is flush left, the second centered, the third flush right.
9.1.4
Aligned columns Aligned columns, on the other hand, are only of one kind, aligned by you. For instance, f (x) = x + yz
g(x) = x + y + z
h(x) = xy + xz + yz
k(x) = (x + y)(x + z)(y + z)
is coded with the alignat environment. It has two aligned columns, both aligned at the = sign.
9.1.5
Touring the Visual Guide Figure 9.1, the Visual Guide, shows thumbnail pictures of the various kinds of multiline math environments and subsidiary math environments. The first part of the Visual Guide illustrates gather and multline. The gather environment is a onecolumn, centered math environment—discussed in Section 9.2— which is used to display a number of formulas collected into one multiline formula. In contrast, multline—discussed in Section 9.3—displays one long formula in a number of lines. The first line is set flush left, the last line set flush right, and the rest (if any) of the lines are centered. The third part of the Visual Guide illustrates the align environment and two of its variants, alignat and flalign, discussed in Section 9.5.
9.2
211
Gathering formulas
Three adjusted subsidiary math environments—matrix, cases, and array—are illustrated in second part of the Visual Guide and presented in Section 9.7. The aligned subsidiary math environments aligned and gathered look just like the align and gather environments, so they are not illustrated in the Visual Guide. The aligned and gathered environments—along with \itemref—are discussed in Section 9.6, along with the split subsidiary math environment; this last one is illustrated in the last part of the Visual Guide.
9.2
Gathering formulas The gather environment groups a number of oneline formulas, each centered on a separate line: (1)
x1 x2 + x21 x22 + x3 ,
(2)
x1 x3 + x21 x23 + x2 ,
(3)
x1 x2 x3 .
Formulas (1)–(3) are typed as follows: \begin{gather} x_{1} x_{2}+x_{1}^{2} x_{2}^{2} + x_{3},\label{E:1.1}\\ x_{1} x_{3}+x_{1}^{2} x_{3}^{2} + x_{2},\label{E:1.2}\\ x_{1} x_{2} x_{3}.\label{E:1.3} \end{gather}
Rule
gather environment 1. Lines are separated with \\. Do not type a \\ at the end of the last line! 2. Each line is numbered unless it has a \tag or \notag on the line before the line separator \\. 3. No blank lines are permitted within the environment. The gather* environment is like gather, except that all lines are unnumbered. They can still be \tagged. It would seem natural to code formulas (1)–(3) with three equation environments: \begin{equation} x_{1} x_{2}+x_{1}^{2} x_{2}^{2} + x_{3},\label{E:1.1} \end{equation} \begin{equation}\label
212
Chapter 9
Multiline math displays
x_{1} x_{3}+x_{1}^{2} x_{3}^{2} + x_{2},\label{E:1.2} \end{equation} \begin{equation} x_{1} x_{2} x_{3}.\label{E:1.3} \end{equation} Note how bad this looks typeset:
9.3
(1)
x1 x2 + x21 x22 + x3 ,
(2)
x1 x3 + x21 x23 + x2 ,
(3)
x1 x2 x3 .
Splitting long formulas The multline environment is used to split one very long formula into several lines. The first line is set flush left, the last line is set flush right, and the middle lines are centered: (4)
(x1 x2 x3 x4 x5 x6 )2 + (y1 y2 y3 y4 y5 + y1 y3 y4 y5 y6 + y1 y2 y4 y5 y6 + y1 y2 y3 y5 y6 )2 + (z1 z2 z3 z4 z5 + z1 z3 z4 z5 z6 + z1 z2 z4 z5 z6 + z1 z2 z3 z5 z6 )2 + (u1 u2 u3 u4 + u1 u2 u3 u5 + u1 u2 u4 u5 + u1 u3 u4 u5 )2
This formula is typed as \begin{multline}\label{E:mm2} (x_{1} x_{2} x_{3} x_{4} x_{5} x_{6})^{2}\\ + (y_{1} y_{2} y_{3} y_{4} y_{5} + y_{1} y_{3} y_{4} y_{5} y_{6} + y_{1} y_{2} y_{4} y_{5} y_{6} + y_{1} y_{2} y_{3} y_{5} y_{6})^{2}\\ + (z_{1} z_{2} z_{3} z_{4} z_{5} + z_{1} z_{3} z_{4} z_{5} z_{6} + z_{1} z_{2} z_{4} z_{5} z_{6} + z_{1} z_{2} z_{3} z_{5} z_{6})^{2}\\ + (u_{1} u_{2} u_{3} u_{4} + u_{1} u_{2} u_{3} u_{5} + u_{1} u_{2} u_{4} u_{5} + u_{1} u_{3} u_{4} u_{5})^{2} \end{multline}
9.3
Splitting long formulas
Rule
213
multline environment 1. Lines are separated with \\. Do not type a \\ at the end of the last line! 2. The formula is numbered as a whole unless it is \tagged or the numbering is suppressed with \notag. (Alternatively, use the multline* environment.) 3. No blank lines are permitted within the environment. 4. Each line is a subformula (see Section 9.4.2). If you are very observant, you may have noticed that we failed to type {}+ following the line separators of the formula. In Section 8.1.2, you were told that this omission would result in the second line being typeset as +(y1 y2 y3 y4 y5 + y1 y3 y4 y5 y6 + y1 y2 y4 y5 y6 + y1 y2 y3 y5 y6 )2 The multline environment, however, knows that a long formula is being broken and so typesets + as a binary operation. A common mistake is to write multiline for multline, resulting in the message: ! LaTeX Error: Environment multiline undefined. In the multline* environment, the formula is not numbered but can be \tagged. The indentation of the first and last lines is controlled by the \multlinegap length command, with a default of 10 points, unless there is a tag on one of those lines. You can adjust the indentation by enclosing the multline environment in a setlength environment (see Section 15.5.2 ), as follows: \begin{multline*} (x_{1} x_{2} x_{3} x_{4} x_{5} x_{6})^{2}\\ + (x_{1} x_{2} x_{3} x_{4} x_{5} + x_{1} x_{3} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{3} x_{5} x_{6})^{2}\\ + (x_{1} x_{2} x_{3} x_{4} + x_{1} x_{2} x_{3} x_{5} + x_{1} x_{2} x_{4} x_{5} + x_{1} x_{3} x_{4})^{2} \end{multline*} \begin{setlength}{\multlinegap}{0pt} \begin{multline*} (x_{1} x_{2} x_{3} x_{4} x_{5} x_{6})^{2}\\ + (x_{1} x_{2} x_{3} x_{4} x_{5} + x_{1} x_{3} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{4} x_{5} x_{6}
214
Chapter 9
Multiline math displays
+ x_{1} x_{2} x_{3} x_{5} x_{6})^{2}\\ + (x_{1} x_{2} x_{3} x_{4} + x_{1} x_{2} x_{3} x_{5} + x_{1} x_{2} x_{4} x_{5} + x_{1} x_{3} x_{4})^{2} \end{multline*} \end{setlength} which typesets as (x1 x2 x3 x4 x5 x6 )2 + (x1 x2 x3 x4 x5 + x1 x3 x4 x5 x6 + x1 x2 x4 x5 x6 + x1 x2 x3 x5 x6 )2 + (x1 x2 x3 x4 + x1 x2 x3 x5 + x1 x2 x4 x5 + x1 x3 x4 )2
(x1 x2 x3 x4 x5 x6 )2 + (x1 x2 x3 x4 x5 + x1 x3 x4 x5 x6 + x1 x2 x4 x5 x6 + x1 x2 x3 x5 x6 )2 + (x1 x2 x3 x4 + x1 x2 x3 x5 + x1 x2 x4 x5 + x1 x3 x4 )2 Notice that the second variant is not indented. Any line of a multline environment can be typeset flush left or right by making it the argument of a \shoveleft or \shoveright command, respectively (same with multline*). For instance, to typeset the second line of formula (4) flush left, as in (x1 x2 x3 x4 x5 x6 )2 + (x1 x2 x3 x4 x5 + x1 x3 x4 x5 x6 + x1 x2 x4 x5 x6 + x1 x2 x3 x5 x6 )2 + (x1 x2 x3 x4 + x1 x2 x3 x5 + x1 x2 x4 x5 + x1 x3 x4 x5 )2 type the formula as follows: \begin{multline*} (x_{1} x_{2} x_{3} x_{4} x_{5} x_{6})^{2}\\ \shoveleft{+ (x_{1} x_{2} x_{3} x_{4} x_{5} + x_{1} x_{3} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{3} x_{5} x_{6})^{2}}\\ + (x_{1} x_{2} x_{3} x_{4} + x_{1} x_{2} x_{3} x_{5} + x_{1} x_{2} x_{4} x_{5} + x_{1} x_{3} x_{4} x_{5})^{2} \end{multline*} Observe that the entire line is the argument of the \shoveleft command, which is followed by \\ unless it is the last line of the environment.
9.4
Some general rules
9.4
215
Some general rules
9.4.1 General rules Even though you have only seen a few examples of multiline math environments, I venture to point out now that the multiline math environments and subsidiary math environments share a number of rules.
Rule
Multiline math environments 1. Lines are separated with \\. Do not type a \\ at the end of the last line! 2. No blank lines are permitted within an environment. 3. If an environment contains more than one formula, then, as a rule, each formula is numbered separately. If you add a \label command to a line, then the equation number generated for that line can be crossreferenced. 4. You can suppress the numbering of a line by using a \notag command on the line. 5. You can also override numbering with the \tag command, which works just as it does for equations (see Section 8.5). 6. \tag and \label should always precede the line separator \\ for lines that are regarded as formulas in their own right. For instance, the lines of the multline environment cannot be individually numbered or tagged. The \tag command works for individual lines, not for the environment as a whole. 7. For crossreferencing, use \label, \ref, and \eqref in the same way you would for an equation (see Section 10.4.2). 8. Each multiline math environment has a *ed form, which suppresses numbering. Individual formulas can still be \tagged. A \notag command placed after the environment is ignored, but a \tag command gives the error message ! Package amsmath Error: \tag not allowed here.
9.4.2 Subformula rules A formula in the multline environment is split into a number of parts by \\ commands; for instance, formula (4) is split into three parts: 1. (x_{1} x_{2} x_{3} x_{4} x_{5} x_{6})^{2}
216
Chapter 9
Multiline math displays
2.
+ (x_{1} x_{2} x_{3} x_{4} x_{5} + x_{1} x_{3} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{4} x_{5} x_{6} + x_{1} x_{2} x_{3} x_{5} x_{6})^{2}
3.
+ (x_{1} x_{2} x_{3} x_{4}+x_{1} x_{2} x_{3} x_{5} + x_{1} x_{2} x_{4} x_{5}+x_{1} x_{3} x_{4} x_{5})^{2}
Such parts of a formula are called subformulas. The first line of the aligned formula r2 = s2 + t2 —from the simple alignment example in Section 3.4.2—which is typed as r^{2} &= s^{2} + t^{2} is split into two parts: 1. 2.
r^{2} = s^{2} + t^{2}
In general, in a line of an aligned formula, the first part is everything between the beginning of the line and the first & symbol. There can then be a number of parts delimited by two consecutive & symbols. Finally, the last part is from the last & symbol to the end of the line or the line separator \\. These parts are also called subformulas. Here are the last of the general rules.
Rule
Subformula 1. Each subformula must be a formula that LATEX can typeset independently. 2. If a subformula starts with the binary operation + or , type it as {}+ or {}. 3. If a subformula ends with the binary operation + or , type it as +{} or {}. Suppose that you want to split the formula x1 + y1 +
X 5 i & a + b + c \end{eqnarray} which typesets as (8)
x =
17y
(9)
y
a+b+c
>
You can type the same formulas with align: \begin{align} x & = 17y\\ y & > a + b + c \end{align} which typesets as (10)
x = 17y
(11)
y >a+b+c
In the eqnarray environment the spacing is based on the spacing of the columns rather than on the spacing requirements of the symbols. I mention eqnarray not for historical reasons but for a very practical one. Unfortunately, a large number of journal submissions still use this construct, and have to be recoded in the editorial offices. Be kind to your editor, and do not use eqnarray.
9.5
223
Aligned columns
9.5.3 The subformula rule revisited Suppose that you want to align the formula x1 + y1 +
X 5 i
with
X 5 i
i
i
!2 + a2 !2
+α
2
so that the + a2 in the first formula aligns with the + α2 in the second formula. You might try typing \begin{align*} x_{1} + y_{1} + \left( \sum_i \binom{5}{i} &+ a^{2} \right)^{2}\\ \left( \sum_i \binom{5}{i} &+ \alpha^{2} \right)^{2} \end{align*} But when you typeset this formula, you get the error message ! Extra }, or forgotten \right. This alignment structure violates the subformula rule because LATEX cannot typeset x_{1} + y_{1} + \left( \sum_i \binom{5}{i} so it is not a subformula. As another simple example, try to align the + in
a+b 2
with the + in x + y:
\begin{align} \binom{a &+ b}{2}\\ x &+ y \end{align} When typesetting this formula, you get the error message ! Missing } inserted. Again, LATEX cannot typeset the subformula \binom{a. To align the two formulas in the first example, add a \phantom command to push the second line to the right: \begin{align*} &x_{1} + y_{1} + \left( \sum_{i < 5} \binom{5}{i} + a^{2} \right)^{2}\\ &\phantom{x_{1} + y_{1} + {}} \left( \sum_{i < 5} \binom{5}{i} + \alpha^{2} \right)^{2} \end{align*}
224
Chapter 9
Multiline math displays
yielding x1 + y1 +
X 5 i\omit
l.12 \end{equation}
\mscount #1\relax \loop \ifnum \mscount ...
The \hdotsfor command also takes an optional argument, a number that multiplies the spacing between the dots. The default is 1. For instance, if we replace \hdotsfor{7} in the previous example by \hdotsfor[3]{7}, then we get (27)
1 2 3 4 5 6 7 8 9 10 1 2 3 . . . . . . . . . . . . .
11 12 11 12
234
Chapter 9
Multiline math displays
We can replace a part of a matrix column with a large figure. 0...0 1 0...0 (a11 ) 0 a = ··· , (ak1 ) = · · · (an1 ) 0
0
0
0
typed as \newcommand{\BigFig}[1]{\parbox{12pt}{\Huge #1}} \newcommand{\BigZero}{\BigFig{0}} \[ a=\left( \begin{matrix} (a_{11})\\ \cdots & \BigZero \\ (a_{n1})\\ \end{matrix} \right) ,\quad (a_{k1})=\left( \begin{matrix} 0\ldots 0 & 1 & 0\ldots 0\\ & 0\\ \BigZero & \cdots & \BigZero\\ & 0\\ \end{matrix} \right) \] Matrix variants A matrix may be enclosed by delimiters (see Section 7.5.1) in a number of different ways: a+b+c uv a+b+c uv a+b+c uv a+b c+d a+b c+d a+b c+d
a + b + c
a + b + c uv uv a+b+c uv
a+b
a+b c + d c + d a+b c+d The first matrix is typed as \begin{matrix} a + b + c & uv\\ a + b & c + d \end{matrix}
9.7
235
Adjusted columns
The others are typed in the same way, except that they use the pmatrix, bmatrix, vmatrix, Vmatrix, and Bmatrix environments, respectively. We can use other delimiters, as in \begin{equation*} \left( \begin{matrix} 1 & 0 0 & 1 \vdots & \vdots 0 & 0 \end{matrix} \right] \end{equation*}
& & & &
\dots \dots \ddots \dots
which produces
& & & &
0\\ 0\\ \vdots\\ 1
1 0 . ..
0 1 .. .
... ... .. .
0 0 .. .
0
0
...
1
This example also uses vertical dots provided by the \vdots commands and diagonal dots provided by the \ddots commands. Small matrix If you put a matrix in an inline mathformula, it may betoo large. Instead, use the a+b+c uv smallmatrix environment. Compare , typed as a+b c+d $\begin{pmatrix} a + b + c & uv\\ a + b & c + d \end{pmatrix}$ with the small matrix
a+b+c uv a+b c+d
, typed as
$\left( \begin{smallmatrix} a + b + c & uv\\ a + b & c + d \end{smallmatrix} \right)$ There are no delimited variants of smallmatrix similar to those of matrix. Instead, use the \left and \right commands with delimiters to enclose a small matrix. The \hdotsfor command does not work in a small matrix.
236
9.7.2
Chapter 9
Multiline math displays
Arrays The array subsidiary math environment is similar to the matrix subsidiary math environment. There are two major differences, however. For an array you must specify the alignment of each column and you have more options to customize it. The first matrix in the introduction to Section 9.7 would be typed as follows using the array subsidiary math environment: \begin{equation*} \left( \begin{array}{cccc} a + b + c & uv & x  y & 27\\ a + b & u + v & z & 134 \end{array} \right) \end{equation*} which produces
Rule
a+b+c uv a+b u+v
x−y z
27 134
array subsidiary math environment 1. Adjacent columns are separated by an ampersand (&). 2. The argument of \begin{array} is mandatory. The argument is a series of the letters l, r, or c, signifying that the corresponding column in the array should be set flush left, flush right, or centered, respectively. The matrix
a+b+c uv a+b u+v
x−y z
27 134
could not have been typeset with matrix since the last column is set flush right. Of course, this is not quite true. In a matrix environment, \hfill 27 would force the number 27 to be set flush right (see Section 5.8.4). If the argument of \begin{array} is missing, as in \begin{equation} \begin{array} a + b + c & uv a + b & u + v \end{array} \end{equation}
& x  y & 27\\ & z & 134
9.7
237
Adjusted columns LATEX generates the error message ! Package array Error:
Illegal preamtoken (a): ‘c’ used.
If you change the first entry of the matrix to c + b + a, then the error message is ! Extra alignment tab has been changed to \cr. \endtemplate l.5
c + b + a &
uv
& x  y & 27\\
Note that the first character in c + b + a is not an Illegal character in array arg. because c is one possible argument of \begin{array}. If the closing brace of the argument of \begin{array} is missing, as in \begin{equation} \begin{array}{cccc a + b + c & uv & x  y & 27\\ a + b & u + v & z & 134 \end{array} \end{equation} you get the error message Runaway argument? {cccc a + b + c & uv
& x  y & 27\\ a + b & u + v \ETC. ! Paragraph ended before \@array was complete. In fact, the argument of array can be more complex than stated in the rule. Indeed, the array subsidiary math environment can take any argument that the tabular environment can take (see Section 6.6). For instance, here is a matrix with headers: 1 2 2
a 1 1 2
b c 1 1 −1 −1 1 0
238
Chapter 9
Multiline math displays
typed as \[ \begin{array}{rrrr} & a & b & c \\ \hline 1 & 1 & 1 & 1 \\ 2 & 1 & 1 & 1 \\ 2 & 2 & 1 & 0 \end{array} \] In Section 9.7.1 we have the matrix example:
(a11 )
a = ···
(an1 )
0
If rows are spanned, we need to use array instead of matrix: a c
b d
0
0
m k
n l
typed as (the \BigZero command is defined on page 234) \left[ \hspace{\arraycolsep} % spacing is automatic with matrix but not with array \begin{array}{cccc} a & b &\multicolumn{2}{c}{}\\ c & d &\multicolumn{2}{c} {\raisebox{1.5ex}[0pt]{\BigZero}}\\ \multicolumn{2}{c}{}& m & n \\ \multicolumn{2}{c} {\raisebox{1.5ex}[0pt]{\BigZero}}& k & l \end{array} \hspace{\arraycolsep} \right] \end{equation*}
9.7
239
Adjusted columns
9.7.3 Cases The cases environment is also a subsidiary math environment. Here is the example from Section 3.4.3 and the introduction to this section: if x < 0; −x2 , f (x) = α + x, if 0 ≤ x ≤ 1; x 2 , otherwise. It is typed as \begin{equation} f(x)= \begin{cases} x^{2}, \alpha + x, x^{2}, \end{cases} \end{equation}
&\text{if $x < 0$;}\\ &\text{if $0 \leq x \leq 1$;}\\ &\text{otherwise.}
It would be easy to code the cases environment as a special case of the array subsidiary math environment: \begin{equation} f(x) = \left\{ \begin{array}{ll} x^{2}, &\text{if $x < 0$;}\\ \alpha + x, &\text{if $0 \leq x \leq 1$;}\\ x^{2}, &\text{otherwise.} \end{array} \right. \end{equation} or of the alignedat subsidiary math environment: \begin{equation*} f(x) = \left\{ \begin{alignedat}{2} &x^{2}, &&\quad\text{if $x < 0$;}\\ &\alpha + x, &&\quad\text{if $0 \leq x \leq 1$;}\\ &x^{2}, &&\quad\text{otherwise.} \end{alignedat} \right. \end{equation*}
240
9.8
Chapter 9
Multiline math displays
Commutative diagrams The amscd package provides the CD subsidiary math environment for typesetting simple commutative diagrams. To use it, make sure that the command \usepackage{amscd} is in the preamble of the document. For instance, to obtain
type \[
\]
\begin{CD} A @>>> @VVV C @= \end{CD}
A −−−−→ y
B y
C
D
B\\ @VVV\\ D
A commutative diagram is a matrix made up of two kinds of rows, horizontal rows, that is, rows with horizontal arrows; and vertical rows, rows with vertical arrows. For example, A
@>>>
B
is a typical horizontal row. It defines two columns and a connecting horizontal arrow @>>>. There may also be more than two columns, as in A
@>>>
B
@>>>
C
@=
D
@
\mathbb{C} @>H_{2}>>\mathbb{C}\\ @VP_{c,3}VV @VP_{\bar{c},3}VV @VVP_{c,3}V\\ \mathbb{C} @>H_{1}>> \mathbb{C} @>H_{2}>> \mathbb{C} \end{CD} Here is another example utilizing the \text command, followed by its source: log
A −−−−→ B −−−−→ bottom oneoney X x β
Y −−−−→ x γ α
C x onto
D ←−−−− E ←−−−− F
Z −−−−→ y
U y
D −−−−→ E −−−−→ H
I
242
Chapter 9 \[
\]
\begin{CD} A @>\log>> B @= D @> @A\beta AA D @>\alpha>> @. \end{CD}
@>>\text{bottom}> @
Z U\\ @AA\gamma A @VVV @VVV\\ E @>>> H I\\
Diagrams requiring more advanced commands should be done with a drawing (or drafting) application or with specialized packages. The AMS recommends Kristoffer H. Rose and Ross Moore’s xypic package (see Section E.1).
9.9
Adjusting the display By default, the math environments described in this chapter do not allow page breaks. While a page break in a cases environment is obviously not desirable, it may be acceptable in an align or gather environment. You can allow page breaks by using the \allowdisplaybreaks command. It allows page breaks in a multiline math environment within its scope. For instance, {\allowdisplaybreaks \begin{align}\label{E:mm13} a &= b + c,\\ d &= e + f,\\ x &= y + z,\\ u &= v + w. \end{align} }% end of \allowdisplaybreaks allows a page break after any one of the first three lines. Within the scope of an \allowdisplaybreaks command, use the \\* command to prohibit a break after that line. The line separators \\ and \\* can use an optional argument to add some additional interline space (see Section 5.7.2). Just before the line separator command (\\), include a \displaybreak command to force a break, or a
9.9
Adjusting the display
243
\displaybreak[0] command to allow one. \displaybreak[n ], where n is 1, 2, or 3, specifies the intermediate steps between allowing and forcing a break. \displaybreak[4] is the same as \displaybreak. You can easily visualize these rules: allow display break = \displaybreak[0] \displaybreak[1] ... \displaybreak[4] = \displaybreak = force display break Note the similarity between the displaybreak sequence and the pagebreak sequence in Section 5.7.3. If you want to allow page breaks in all multiline math environments in your document, place the \allowdisplaybreaks[1] command in the preamble of your document. The optional argument can be varied from 1 to 4, in order of increasing permissiveness. Note that none of the subsidiary math environments are affected by any variant of the \displaybreak or the \allowdisplaybreaks commands.
CHAPTER
10 LATEX documents
In this chapter, we take up the organization of shorter documents. Longer documents and books are discussed in Part VI. If you are writing a simple article, start with a template (see Sections 4.2 and 11.4), then you can safely ignore much of the material discussed in this chapter. In more complicated articles you may need the material discussed in this chapter. Section 10.1 discusses document structure in general, Section 10.2 presents the preamble. Section 10.3 discusses the top matter, in particular, the abstract environment. Section 10.4 presents the main matter, including sectioning, crossreferencing, tables, and figures. Section 10.5 covers the back matter, including the bibliography and index. In Section 10.1–10.5 we discuss the logical design of a LATEX document. The visual design is largely left to the document class. In Section 10.6, however, we briefly discuss one frequently adjusted aspect of visual design, the page style.
248
Chapter 10
\documentclass{...} \usepackage{...} ...
LATEX documents
preamble
\begin{document}
top matter
\title{...} \author{...} \address{...} \date{...} \begin{abstract} ... \end{abstract} \maketitle
main matter
abstract
body
\section{...} \section{...}
\begin{thebibliography}{9} back matter ... \end{thebibliography} \end{document}
Figure 10.1: The structure of a LATEX document.
10.1
The structure of a document The source file of a LATEX document is divided into two main parts: the preamble and the body (see Figure 10.1). Preamble This is the portion of the source file before the \begin{document} command. It contains definitions and instructions that affect the entire document. Body This is the content of the document environment. It contains all the material to be typeset. These statements oversimplify the situation somewhat. For instance, you can define a command in the preamble to typeset some text that will appear wherever the command is used in the body, but the text is actually typed in the preamble. Nevertheless, I hope the division between the preamble and the body is clear.
10.2
The preamble
249
The body is divided into three parts: Top matter This is the first part of the body. It is concluded with the \maketitle command. Traditionally it included only the \title, the \author, and the \date commands. The top matter is derived from these commands and from it the title page of an article was designed. This evolved to include a lot more information about the author(s), for instance, their email addresses, academic affiliations, home pages, and about the article, for instance, research support, subject classification. The typeset top matter now is split into several locations, the top and bottom of the first page and the bottom of the last page. See pages 286 and 288 for an example and Section 18.1.2 for more components that can be used in longer documents and books. Main matter This is the main part of the document, including any appendices. Back matter This is the material that is typeset at the end of the document. For a typical shorter document, the back matter is just the bibliography. See Section 18.1.2 for more information about additional components—such as the index—that are often used in longer documents and books.
10.2
The preamble You were introduced to the preamble of a document in Section 4.1. Recall that the preamble contains the crucial \documentclass line, specifying the document class and the options that modify its behavior. For instance, \documentclass[draft,reqno]{amsart} loads the document class amsart with the draft option, which paints a slug in the margin indicating lines that are too wide (see Section 5.7.1), and the reqno option, which places the equation numbers on the right (see Section 11.5). article is the most popular legacy document class (see Section 12.1). The command \documentclass[titlepage,twoside]{article} loads the document class article with the titlepage option, which creates a separate title page and places the abstract on a separate page, and the twoside option, which formats the typeset article for printing on both sides of the paper. The \documentclass command is usually followed by the \usepackage commands, which load LATEX enhancements called packages. For instance, \usepackage{latexsym} loads a package that defines some additional LATEX symbol names (see Section 12.3), whereas
250
Chapter 10
LATEX documents
\usepackage[demo]{graphicx} loads the graphicx package (see Section 10.4.3) with the demo option that inserts rectangles in place of the illustrations. Document class options are also passed on to the packages as possible options, so \documentclass[demo]{amsart} \usepackage{graphicx} would also load the graphicx package with the demo option unless it is invoked with \usepackage[final]{graphicx} Any document class options that are not relevant for a package are ignored. \usepackage commands can also be combined: \usepackage{amssymb,latexsym} is the same as \usepackage{amssymb} \usepackage{latexsym} Document class files have a cls extension, whereas package files are designated by the sty extension. The document class amsart is defined in the amsart.cls file, the graphicx package is defined in the graphicx.sty file. You may define your own packages, such as the newlattice package described in Section 15.3. The preamble normally contains any userdefined commands (see Chapter 15) and the proclamation definitions (see Section 6.4). Some commands can only be in the preamble. \DeclareMathOperator is such a command (see Section 7.6.2) and so is \numberwithin (see Section 7.3). If you put such a command in the body, for example, \DeclareMathOperator, you get an error message: ! LaTeX Error: Can be used only in preamble. l.103 \DeclareMathOperator There is one command that may only be placed before the \documentclass{...} line: \NeedsTeXFormat{LaTeX2e}[2005/12/01] This command checks the version of LATEX being used to typeset the document and issues a warning if it is older than December 1, 2005 or whatever date you specified. Use this optional date argument if your document contains a feature that was introduced on or after the date specified or if an earlier version had a bug that would materially affect the typesetting of your document.
10.3
Top matter
251
For instance, if you use the \textsubscript command, introduced in the December 1, 2005 release (see page 311), then you may use the \NeedsTeXFormat line shown above. LATEX now hardly changes from year to year, so this command is rarely used except in document class files or package files. See, however, the discussion on page 311.
10.3
Top matter The top matter of an article is part of the article body and, as a rule, it contains the material used to create the “title page” and, optionally, an abstract. Discussion of the top matter should take place in the context of a particular document class. We discuss the top matter of the amsart document class in Section 4.2, and we continue discussing it in much more detail in Section 11.2. The top matter of the article document class is covered in Section 12.1.1. Long documents, such as books, have rather complicated top matter such as tables of contents (see Chapter 18). In this section, we only discuss the abstract.
10.3.1 Abstract Most standard document classes, except those for letters and books, make provision for an abstract, typed in an abstract environment. The document class formats the heading as Abstract, or some variant, and, as a rule, typesets the text of the abstract in smaller type with wider margins. The amsart document class requires that you place the abstract environment before the \maketitle command (see Figure 10.1). See the abstract in the sample article intrart.tex on page 37. If you forget to place it there, you get the warning Class amsart Warning: Abstract should precede \maketitle in AMS documentclasses; reported on input line 21. and the abstract is typeset wherever the abstract environment happens to be placed. In the article document class you place the abstract after the \maketitle command. If you place the abstract before the \maketitle command, the abstract is placed on page 1, and the article starts on page 2. If the abstract and the “footnotes” from the top matter fill the first page, the second page has no running head. To fix this, follow the \maketitle command with the \clearpage command (see Section 5.7.3).
10.4
Main matter The main matter contains most of the essential parts of the document, including the appendices.
252
Chapter 10
LATEX documents
We discuss now how to structure the main matter. We describe sectioning in Section 10.4.1, crossreferencing in Section 10.4.2, and tables and figures in Section 10.4.3.
10.4.1
Sectioning The main matter of a typical shorter document is divided into sections. See Section 18.1.1 for a discussion on sectioning longer documents. Sections LATEX is instructed to start a section with the \section command, which takes the title of the section as its argument. This argument may also be used for the running head and it is also placed in the table of contents (see Section 18.2), which means that you need to protect fragile commands with the \protect command (see Section 5.3.3). LATEX automatically assigns a section number and typesets the section number followed by the section title. Any \section command may be followed by a \label command, so that you can refer to the section number generated by LATEX, as in \section{Introduction}\label{S:intro} The command \ref{S:intro} refers to the number of the section and the command \pageref{S:intro} refers to the number of the typeset page where the section title appears. Other sectioning commands A section may be subdivided into subsections, which may themselves be divided into subsubsections, paragraphs, and subparagraphs. Subsections are numbered within a section (in Section 1, they are numbered 1.1, 1.2, and so on). Here is the whole hierarchy: \section \subsection \subsubsection \paragraph \subparagraph It is important to understand that the five levels of sectioning are not just five different styles for typesetting section headers but they form a hierarchy. You should never have a subsection outside a section, a subsubsection outside a subsection, and so on. For instance, if the first sectioning command in your document is \subsection, the subsections are numbered 0.1, 0.2, . . . . Or if in the first section of your document
10.4
Main matter
253
the first sectioning command is \subsubsection, the subsubsections are numbered 1.0.1, 1.0.2, . . . . Both are clearly undesirable. There are two additional sectioning commands provided by the report and by the book document classes (book and amsbook): \chapter and \part (see Section 18.1.1). Any sectioning command may be followed by a \label command so that you can refer to the number (if any) generated by LATEX and the page on which it appears (see Section 10.4.2). There is also the seldom used top level \specialsection command. Articles do not have parts and chapters, but sometimes a long article may require further division using the \specialsection command. The form of sectioning commands All sectioning commands take one of the following three forms, illustrated below with the \section command: Form 1 The simplest form is \section{title } where title is the section title, of course. You need to protect any fragile commands in title with the \protect command (see Section 5.3.3). Form 2 The sectioning command may have an optional argument \section[short_title ]{title } The optional short title argument is used in the running head. See Section 18.2 on what goes into the table of contents. Protect any fragile commands in short title with the \protect command (see Section 5.3.3). Form 3 Finally, we consider the *ed version \section*{title } There are no section numbers printed and the title is not included in the running head. Remember that if you * a section, all subsections, and so on, must also be *ed to avoid having strange section numbers. Sectioning commands typeset Consider the following text: \section{Introduction}\label{S:Intro} We shall discuss the main contributors of this era. \subsection{Birkhoff’s contributions}\label{SS:contrib}
254
Chapter 10
LATEX documents
\subsubsection{The years 19351945}\label{SSS:1935} Going to Oxford was a major step. \paragraph{The first paper} What should be the definition of a universal algebra? \subparagraph{The idea} One should read Whitehead very carefully. This is how it looks typeset in the amsart document class:
1
Introduction
We shall discuss the main contributors of this era.
1.1 1.1.1
Birkhoff ’s contributions The years 1935–1945
Going to Oxford was a major step. The first paper What should be the definition of a universal algebra? The idea
One should read Whitehead very carefully.
Notice that paragraphs and subparagraphs are not displayed prominently by the AMS. By contrast, look at the same text typeset BLAHin the legacy article document class:
1. Introduction We shall discuss the main contributors of this era. 1.1. Birkhoff ’s contributions. 1.1.1. The years 1935–1945. Going to Oxford was a major step. The first paper. What should be the definition of a universal algebra? The idea. One should read Whitehead very carefully.
This illustrates vividly one huge difference between the two document classes, the visual handling of sectioning. Section 15.5.1 discusses how you can change the format of the section numbers, and how to specify which sectioning levels are to be numbered.
1
10.4
Main matter
255
Section 2.2 of The LATEX Companion, 2nd edition [46] explains how to change the layout of section headings, especially useful for document class designers. Appendix In the main matter, if the article contains appendices, mark the beginning of the appendices with the \appendix command. After the \appendix command, the \section command starts the appendices (for books, see Section 18.1.2): \appendix \section{A proof of the Main Theorem}\label{S:geom} This produces Appendix A with the given title, typeset just like a section. Note that appendices may be labeled and crossreferenced like any other section. In an appendix, subsections are numbered A.1, A.2, and so on, subsubsections within A.1 are numbered A.1.1, A.1.2, and so on. Let me repeat, \appendix is not like \section. It is not a command with an argument. Appendices are named by arguments of the \section—or for books, \chapter—commands placed after the \appendix command.
10.4.2 Crossreferencing There are three types of crossreferencing available in LATEX: 1. Symbolic referencing with \ref and \eqref for equations 2. Page referencing with \pageref 3. Bibliographic referencing with \cite In this section, we discuss the first two, while bibliographies are discussed in Section 10.5.1 and in Chapter 16. Symbolic referencing Wherever LATEX can automatically generate a number in your document, you can place a \label command \label{symbol } Then, at any place in your document, you can use the \ref command \ref{symbol } to place that number in the document. We call symbol the label. You can use labels for sectioning units, equations, figures, tables, items in an enumerated list environment (see Section 6.2.1), as well as for theorems and other proclamations.
256
Chapter 10
LATEX documents
If the equation labeled E:int is the fifth equation in an article, then LATEX stores the number 5 for the label E:int, so \ref{E:int} produces the number 5. If equations are numbered within sections (see Section 7.3), and an equation is the third equation in Section 2, then LATEX stores the number 2.3 for the label E:int, so the reference \ref{E:int} produces the number 2.3. Example 1 The present section starts with the command \section{Main matter}\label{S:MainMatter} So \ref{S:MainMatter} produces the number 10.4 and we get the number of the typeset page where the section title appears with \pageref{S:MainMatter}, which is 251. Example 2 \begin{equation}\label{E:int} \int_{0}^{\pi} \sin x \, dx = 2. \end{equation} In this case, \ref{E:int} produces the number of the equation, \eqref{E:int} produces the number of the equation in parentheses.
Tip
If you have to reference an equation in the statement of a theorem, always use \eqref. Do not use \eqref to reference anything but proclamations. (See the \itemref command introduced in Section 15.1.2.)
Example 3 \begin{theorem}\label{T:fund} Statement of theorem. \end{theorem} The reference \ref{T:fund} produces the number of the theorem.
Tip
Typeset a document twice to see a change in a crossreference. See Section D.3.4 for a discussion of how LATEX stores these numbers and why you have to typeset twice. If you typeset only once, and LATEX suspects that the crossreferences have not been updated, you get a warning: LaTeX Warning: Label(s) may have changed. Rerun to get crossreferences right.
10.4
Main matter
257
Rule 1
\label command The argument of the \label command is a string of letters, punctuation marks, and digits. It is case sensitive, so S:intro is different from S:Intro.
Rule 2
\label command Place a \label command immediately after the command that generates the number. The following is not compulsory but advisable.
Tip
When referencing: see Section~\ref{S:Intro} proved in Theorem~\ref{T:main} or see Sections~\ref{S:Intro} and~\ref{S:main} use ties (~). It is difficult to overemphasize how useful automatic crossreferencing can be when writing a document.
Tip
Make your labels meaningful to yourself, so they are easy to remember. Systematize your labels. For example, start the label for a section with S:, theorem with T:, lemma with L:, and so on. When you are crossreferencing, even if you follow these tips, it may not be easy to remember a label. David Carlisle’s showkeys package may help you out. It is part of the tools distribution (see Section 12.3.1 and Section E.1). Include the line \usepackage{showkeys} in the preamble of your document. The showkeys package shows all symbolic references in the margin of the typeset document. With the notcite option, my preference, \usepackage[notcite]{showkeys} showkeys does not show the labels for bibliographic references. When the document is ready for final typesetting, then comment out this line.
258
Chapter 10
LATEX documents
Section 2.4 of The LATEX Companion, 2nd edition [46] describes varioref, a package which extends the power of \ref, and xr, a package for referencing external documents. Page referencing The command \pageref{symbol } produces the number of the typeset page corresponding to the location of the command \label{symbol }. For example, if the following text is typeset on page 5, There may be three types of problems with the construction of such lattices.\label{problem} and you type Because of the problems associated with the construction (see page~\pageref{problem}) anywhere in the document, LATEX produces Because of the problems associated with the construction (see page 5) Because of the way LATEX typesets a page, page references may be off by one. See the discussion in Section 18.6 on how to guarantee that the page number is correct.
10.4.3
Floating tables and illustrations Many documents contain tables and illustrations. These must be treated in a special way since they cannot be broken across pages. If necessary, LATEX moves—floats— a table or an illustration to the top or bottom of the current or the next page if possible and further away if not. LATEX provides the table and the figure environments for typesetting floats. The two are essentially identical except that the figure environments are named Figure 1, Figure 2, and so on, whereas the table environments are numbered as Table 1, Table 2, and so on. Tables A table environment is set up as follows: \begin{table} Place the table here \caption{title }\label{Ta:xxx } \end{table}
10.4
Main matter
259
The \caption command is optional and may also precede the table. The optional \label command must be placed between the command \caption and the command \end{table}. The label is used to reference the table’s number. A table environment can have more than one table, each with its own caption. The table environment is primarily used for tables made with the tabular or similar environments (see Section 6.6). There are many examples of tables in this book, for instance, Section 5.4 has four. If your document uses the twocolumn document class option, the table environment produces tables that span only one column and the table* environment produces tables that span both columns. Such tables can be placed only at the top of a page. Figures Illustrations, also called graphics or figures, include drawings, scanned images, digitized photos, and so on. These can be inserted with a figure environment: \begin{figure} Place the graphics here \caption{title }\label{Fi:xxx } \end{figure} The above discussion of captions and labels for tables also applies to figures. Like the table environment, if your document uses the twocolumn document class option, the figure environment produces figures that span only one column, but the figure* environment produces figures that span both columns. However, these figures can be placed only at the top of a page. The standard way of including a graphics file is with the commands provided by the graphicx package by David Carlisle and Sebastian Rahtz, which is part of the LATEX distribution (see Section 12.3). Save your graphics in EPS (Encapsulated PostScript) or PDF (Portable Document Format) format. Your graphics can also be made within a picture environment, an approach that is neither encouraged nor discussed in this book. Using the graphicx package, a typical figure is specified as follows: \begin{figure} \centering\includegraphics{file } \caption{title }\label{Fi:xxx } \end{figure} The illustration circle.eps is included with the command \includegraphics{circle} without the extension! LATEX and the graphicx package assumes the eps extension. On the other hand, versions of LATEX that produce a typeset file in PDF format assume the pdf extension.
260
Chapter 10
LATEX documents
If you have to scale the graphics image, say to 68% of its original size, use the command \includegraphics[scale=.68]{file } For instance, the figure on page 526 is included with the commands \begin{figure} \centering\includegraphics[scale=.8]{StrucLaT} \caption{The structure of \protect\LaTeX.} \label{Fi:StrucLaT} \end{figure} Float control The table and figure environments may have an optional argument, with which you can influence LATEX’s placement of the typeset table. The optional argument consists of one to four letters: b, the bottom of the page h, here (where the environment appears in the text) t, the top of the page p, a separate page For instance, \begin{table}[ht] requests LATEX to place the table “here” or at the “top” of a page. The default is [tbp] and the order of the optional arguments is immaterial, for example, [th] is the same as [ht]. If h is specified, it takes precedence, followed by t and b. LATEX has more than a dozen internal parameters that control a complicated algorithm that determines the placement of tables and figures. If you want to override these parameters for one table or figure only, add an exclamation mark (!) to the optional argument. For instance, [!h] requests that this table or figure be placed where it is in the source file even if this placement violates the rules as set by some of the parameters. For a detailed discussion of the float mechanism, see Chapter 6 of The LATEX Companion, 2nd edition [46]. The \suppressfloats command stops LATEX from placing any more tables or figures on the page it appears on. An optional argument t or b (but not both) prohibits placement of floats at the top or bottom of the current page. The table or figure that is suppressed appears on the next page or later in the document, if necessary. Your demands and LATEX’s float mechanism may conflict with one another with the result that LATEX may not place material where you want it. The default values
10.5
Back matter
261
of the float placement parameters are good only for documents with a small number of floating objects. Combining two tables or illustrations into one sometimes helps. The \clearpage command not only starts a new page with the \newpage command, but also forces LATEX to print all the tables and figures it has accumulated but not yet placed in the typeset document. See also some related commands discussed in Section 5.7.3. For more information on graphics, see Chapter 10 of The LATEX Companion, 2nd edition [46] and Chapter 2 of The LATEX Graphics Companion [17]. See also the documentation for the graphicx package in the LATEX distribution (see Section 12.3).
10.5
Back matter The back matter of an article is very simple, as a rule. It is either empty or consists of only a bibliography. A long document, such as a book, may have more complicated back matter (see Chapter 18). In this section, we discuss only the bibliography and a very simple index.
10.5.1 Bibliographies in articles The simplest way to typeset a bibliography is to type it directly into the article. For an example, see the bibliography in the intrart.tex article (on page 43). This sample bibliography contains two examples, one short and one long, of each of the seven most frequently used kinds of items. You type the text of a bibliography in a thebibliography environment, as shown in the following examples. \begin{thebibliography}{99} \bibitem{hA70} Henry~H. Albert, \emph{Free torsoids}, Current trends in lattice theory. D.~Van Nostrand, 1970. \bibitem{hA70a} Henry~H. Albert, \emph{Free torsoids}, Current trends in lattice theory (G.\,H. Birnbaum, ed.). vol.~7, D.~Van Nostrand, Princeton, January, 1970, no translation available, pp.~173215 (German). \bibitem{sF90} SooKey Foo, \emph{Lattice Constructions}, Ph.D. thesis, University of Winnebago, 1990.
262
Chapter 10
LATEX documents
\bibitem{sF90a} SooKey Foo, \emph{Lattice Constructions}, Ph.D. thesis, University of Winnebago, Winnebago, MN, December 1990, final revision not yet available. \bibitem{gF86} Grant~H. Foster, \emph{Computational complexity in lattice theory}, tech. report, Carnegie Mellon University, 1986. \bibitem{gF86a} Grant~H. Foster, \emph{Computational complexity in lattice theory}, Research Note 128A, Carnegie Mellon University, Pittsburgh, PA, December, 1986, research article in preparation. \bibitem{pK69} Peter Konig, \emph{Composition of functions}. Proceedings of the Conference on Universal Algebra (Kingston, 1969). \bibitem{pK69a} Peter Konig, \emph{Composition of functions}. Proceedings of the Conference on Universal Algebra (G.~H. Birnbaum, ed.). vol.~7, Canadian Mathematical Society, Queen’s Univ., Kingston, ON, available from the Montreal office, pp.~1106 (English). \bibitem{wL75} William~A. Landau, \emph{Representations of complete lattices}, Abstract: Notices Amer. Math. Soc. \textbf{18}, 937. \bibitem{wL75a} William~A. Landau, \emph{Representations of complete lattices}, Abstract: Notices Amer. Math. Soc. \textbf{18}, 937, December, 1975. \bibitem{gM68} George~A. Menuhin, \emph{Universal algebra}. D.~Van Nostrand, Princeton, 1968.
10.5
Back matter
263
\bibitem{gM68a} George~A. Menuhin, \emph{Universal algebra}. 2nd ed., University Series in Higher Mathematics, vol.~58, D.~Van Nostrand, Princeton, March, 1968 (English), no Russian translation. \bibitem{eM57} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \textbf{8}~(1957), 455460. \bibitem{eM57a} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \textbf{8}~(1957), 455460 (English), Russian translation available. \end{thebibliography} Figure 10.2 shows a typeset version of this bibliography in the amsart document class. By contrast, look at the same bibliography typeset in the legacy article document class in Figure 10.3. You can find these entries in the document inbibl.tpl in the samples folder (see page 4). I use the convention that the label for a \bibitem consists of the initials of the author and the year of publication. The first cited publication by Andrew B. Reich in 1987 would have the label aR87 and the second, aR87a. Of course, you can use any label you choose, but such conventions make the items easier to reuse. The thebibliography environment takes an argument—in the previous example, this argument is 99—telling LATEX that the widest reference number it must generate is two digits wide. For fewer than 10 items, use 9 and for 100 or more items, use 999. If the argument of \begin{thebibliography} is missing, you get the error message ! LaTeX Error: Something’s wrongperhaps a missing \item. Each bibliographic item is introduced with \bibitem, which is used like the \label command. In your text, use \cite, in a similar way to \eqref—it provides the number enclosed in brackets. So if the 13th bibliographic item is introduced with \bibitem{eM57}
264
Chapter 10
LATEX documents
then \cite{eM57} refers to that item and typesets it as [13]. The bibliography of the article itself is automatically numbered by LATEX. It is up to the author to make sure that the listing of the bibliographic items is in the proper order.
Tip
Do not leave spaces in a \cite command. For example, \cite{eM57 } produces [?] indicating an unknown reference. You can use \cite to cite two or more items in the form \cite{hA70,eM57} which typesets as [1, 13]. There is also an optional argument for \cite to specify additional information. For example,
References [1] Henry H. Albert, Free torsoids, Current trends in lattice theory. D. Van Nostrand, 1970. [2] Henry H. Albert, Free torsoids, Current trends in lattice theory (G. H. Birnbaum, ed.). vol. 7, D. Van Nostrand, Princeton, January, 1970, no translation available, pp. 173–215 (German). [3] SooKey Foo, Lattice Constructions, Ph.D. thesis, University of Winnebago, 1990. [4] SooKey Foo, Lattice Constructions, Ph.D. thesis, University of Winnebago, Winnebago, MN, December 1990, final revision not yet available. [5] Grant H. Foster, Computational complexity in lattice theory, tech. report, Carnegie Mellon University, 1986. [6] Grant H. Foster, Computational complexity in lattice theory, Research Note 128A, Carnegie Mellon University, Pittsburgh, PA, December, 1986, research article in preparation. [7] Peter Konig, Composition of functions. Proceedings of the Conference on Universal Algebra (Kingston, 1969). [8] Peter Konig, Composition of functions. Proceedings of the Conference on Universal Algebra (G. H. Birnbaum, ed.). vol. 7, Canadian Mathematical Society, Queen’s Univ., Kingston, ON, available from the Montreal office, pp. 1–106 (English). [9] William A. Landau, Representations of complete lattices, Abstract: Notices Amer. Math. Soc. 18, 937. [10] William A. Landau, Representations of complete lattices, Abstract: Notices Amer. Math. Soc. 18, 937, December, 1975. [11] George A. Menuhin, Universal algebra. D. Van Nostrand, Princeton, 1968. [12] George A. Menuhin, Universal algebra. 2nd ed., University Series in Higher Mathematics, vol. 58, D. Van Nostrand, Princeton, March, 1968 (English), no Russian translation. [13] Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. [14] Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460 (English), Russian translation available.
Figure 10.2: The most important bibliographic entry types.
10.5
265
Back matter
Figure 10.3: Bibliography in the article document class.
266
Chapter 10
LATEX documents
\cite[pages~215]{eM57} typesets as [13, pages 2–15]. If you wish to use labels rather than numbers to identify bibliographic items, then you can specify those labels with an optional argument of the \bibitem command:
typed as \bibitem[EM57]{eM57} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \textbf{8} (1957), 455460. If this optional argument of \bibitem is used, then the \cite command produces [EM57]. The argument of \begin{thebibliography} must be set wide enough to allow for such labels.
Rule
Label for a bibliographic item A label cannot contain a comma or a space. The examples I have used follow the formatting rules set by the AMS. Only titles are italicized, and only volume numbers of journals are set in boldface. You also have to watch the order in which the items are given, the punctuation, and the capitalization. If an author appears repeatedly, use the \bysame command, which replaces the author’s name with a long dash followed by a thin space. For example, \bibitem{gF86} Grant~H. Foster, \emph{Computational complexity in lattice theory}, tech. report, Carnegie Mellon University, 1986. \bibitem{gF86a} \bysame, \emph{Computational complexity in lattice theory}, Research Note 128A, Carnegie Mellon University, Pittsburgh PA, December 1986, research article in preparation. See sampart.tex on page 293 for a typeset example.
10.5
Back matter
Tip
267
If you want a different title for your bibliography, say Bibliography, place the command \renewcommand{\refname}{Bibliography} anywhere before the thebibliography environment (see Section 15.1.6). If you use a legacy document class or amsbook.cls, use the line \renewcommand{\bibname}{Bibliography}
Tip
You may have more than one thebibliography environment in a document. Because each bibliography would number the entries from 1, you should provide labels as optional parameters of the \bibitem commands for crossreferencing.
10.5.2 Simple indexes Using the \label and \pageref commands (see Section 10.4.2), it is quite simple to produce a small index in a theindex environment. At each point in the text that you want to reference in the index, place a \label command. The corresponding entry in the index typesets the page number with the \pageref command. The \item, \subitem, and \subsubitem commands create an entry, subentry, and subsubentry, respectively. If you need additional vertical spacing when the first letter changes, for instance, between the “h” entries and the “i” entries, you can use the \indexspace command. Here are some examples of index entries: \begin{theindex} \item Lakser, H., \pageref{Lakser} \item Lattice, \pageref{Lattice_intro}, \textbf{\pageref{Lattice}} \subitem distributive, \pageref{Lattice_distributive} \subitem modular, \pageref{Lattice_distributive}, \textbf{\pageref{Lattice_distributive2}} \item Linear subspace, \pageref{Linear_subspace} \end{theindex}
268
Chapter 10
LATEX documents
And here is the typeset index:
For a larger index, you should use the MakeIndex application (see Chapter 17).
10.6
Visual design In this chapter, we have discussed the logical design of a LATEX document. The visual design is largely left to the document class. But there is one small aspect of the visual design we have to discuss, the page style. To get a visual representation of the page style of your document, use layout package of Kent McPherson (see Section 12.3.1). Load the package with \usepackage{layout} and place the \layout command somewhere in the body of your article. LATEX produces a graphical representation of the page layout. Figure 10.4 shows the page layout for odd pages for the amsart document class with no options. A typeset page has three parts, the running head or header, the body, and the footer. As a rule, the document class takes care of the contents and formatting of all three parts. For the running head and footer, however, you can override the page design of the document class with the command \pagestyle{style } where the argument style is one of the following: plain The running head is empty and the footer contains only the page number empty Both the running head and the footer are blank headings The running head contains the information provided by the document class and the footer is empty myheadings The running head contains the information provided by the commands \markboth and \markright, the footer is empty The \markright command takes only one argument. The last \markright on a page provides the running head information for that page. The \markboth command has two arguments. The first provides the running head information for a lefthand page, the second provides the running head information for a righthand page. The AMS
10.6
269
Visual design
Figure 10.4: Page layout for the amsart document class.
270
Chapter 10
LATEX documents
document classes also have a \markleft command for the running head information for a righthand page. The \thispagestyle command is the same as \pagestyle except that it affects only the current page. For instance, if the current page is a fullpage graphic, you might want to issue the command \thispagestyle{empty} The \maketitle command automatically issues a \thispagestyle{plain} command, so if you want to suppress the page number on the first page of a document, you have to put \thispagestyle{empty} immediately after the \maketitle command. The commands listed in Figure 10.4 are length commands (see Section 15.5.2) and can be changed with the commands introduced in that section. As a rule, you do not have to worry about these settings, they are chosen by the document class for you. Sometimes, however, you have a job that requires such changes. I once had to submit a research plan on a form with a 7.5 inch by 5 inch box. To be able to cut and paste the typeset report, I had to produce the text with a \textwidth of 7 inches. If I simply set \setlength{\textwidth}{7in} the text would overflow the printed page and the last few characters of each line would be missing. So I had to change the margins by starting the document with \documentclass[12pt]{report} \setlength{\textwidth}{7in} \setlength{\oddsidemargin}{0pt} All of Chapter 4 of The LATEX Companion, 2nd edition [46] deals with page layouts. There you can find a description of the geometry package of Hideo Umeki, which computes all the parameters from the ones you supply. Also you find there a discussion of Piet van Oostrum’s excellent package, fancyhdr, which allows you to create your own page style (see also [24]). See Section E.1 on how to get these packages. However, if you submit an article to a journal, do not change the type size, page dimensions, headers. Use the document class and the article templates the journal provides (if any). This will make your submission easier for you and the journal.
CHAPTER
11 The AMS article document class
In this chapter, we discuss amsart, the main AMS document class for journal articles. The AMS book document class is discussed in Chapter 18. In Section 11.1, I argue that there are good reasons why you should write your articles for publication in amsart. Section 11.2 introduces the rules governing the top matter in the amsart document class. Section 11.3 contains a detailed sample article. We present the source file and the typeset version juxtaposed. In Section 4.2, you created a simple template for amsart articles. In Section 11.4, you are guided through the process for creating more detailed templates. A document class is shaped by its options. In Section 11.5, we discuss the options of amsart. Section 11.6 briefly describes the various packages in the AMS distribution and their interdependencies.
11.1
Why amsart?
11.1.1 Submitting an article to the AMS You want to submit an article written with the amsart document class to the Proceedings of the American Mathematical Society.
272
Chapter 11
The AMS article document class
For general information on the AMS journals, go to the AMS Web site http://www.ams.org/ and start discovering the wealth of relevant information in the column Publications & Tools, especially in the Author Resource Center. Then click on the journal from the list, and on the Proceedings page, click on Author packages. The page that comes up has everything you need, in particular, the procl.cls, the document class for the Proceedings. Click on it to download the document class. Place it in the folder with your article. There is a more direct way to find procl.cls, and this will always work even when the AMS changes the layout of its Web site. At the AMS Web site click on Search. In the search field, type procl.cls. Do the search and click on the link to procl.cls. Now, in the preamble of your article, replace the line \documentclass{amsart} with \documentclass{procl} Typeset the article and you are done. Your article is formatted as it will appear in the Proceedings.
11.1.2
Submitting an article to Algebra Universalis There are many journals whose document class is based on amsart. For instance, http://www.math.umanitoba.ca/homepages/au/ takes you to the home page of the journal Algebra Universalis. To find the document class, click on Instructions for authors and then click where indicated to get au.cls. Now in your article make the replacement \documentclass{au} and your article typesets in the format appropriate for this journal.
11.1.3
Submitting to other journals A large number of journals use document classes based on amsart. Not all are as friendly as Algebra Universalis, but as a rule a small number of changes in the article suffice. All of them share the attribute that the top matter is given as the arguments of many commands. In the introductory sample article, intrart.tex, on page 37, there were only four, but in the sample article sampart.tex, on page 290, there are nine—there
11.2
The top matter
273
could be many more. Contrast this with the legacy article class (see Section 12.1.1). As a result, the document class is able to shape the top matter as the journal requires. Even if the names of some of these commands are different (e.g., affiliation instead of address), the principles you learn from the amsart document class apply. Many journals insist that you use their own document classes. For these, you may have to add the AMS packages (see Section 11.6) to continue to use the enhancements of the AMS. A shrinking number of journals use document classes incompatible with the AMS packages. If you can, avoid these journals.
11.1.4 Submitting to conference proceedings The AMS also has a document class for articles for bookform proceedings of meetings. The differences in the rules for the amsart and amsproc document classes are so minor that you can safely ignore them.
11.2
The top matter For a fairly representative example, see the typeset top matter of the sampart.tex article on pages 286 and 288. As you may recall from Section 4.1, part of the author information is moved to the end of the typeset article—see page 288. Title page information is provided as arguments of several commands. For your convenience, I divide them into three groups: information about the article, information about the author, and AMS related information. There is only one general rule.
Rule
Top matter commands All top matter commands are short. This means that there can be no blank lines (or \par commands) in the argument of any of these commands (see Section 5.3.3).
11.2.1 Article information You have to supply five pieces of information about the article.
Rule
Title Command: \title Separate lines with \\
274
Chapter 11
The AMS article document class
Optional argument: Short title for running head Do not put a period at the end of a title Do not use userdefined commands in the title The typeset title is placed on the front page of the typeset article. Many titles are too long to be typeset on a single line in the font used by the amsart document class for titles. If the way LATEX breaks the title is not satisfactory, you can indicate where the title should be broken with the \\ command. Alternatively, you may nudge LATEX in the right direction with ~ (see Section 5.4.3). For instance, the title: The \texttt{amsart} document class is broken by LATEX between document and class. So either add \\: The \texttt{amsart}\\ document class or replace document class with document~class: The \texttt{amsart} document~class The running head (see Section 10.6) is the title on oddnumbered pages, set in capital letters. If the title is more than a few words long, use an optional argument to specify a short title for the running head. Do not use \\ in the short title. Example of a title: \title{A construction of distributive lattices} A title with a short title: \title[Completesimple distributive lattices] {A construction of\\ completesimple distributive lattices} Note the AMS rules about short titles and the table of contents on Section 18.2.1.
Rule
Translator Command: \translator Do not put a period at the end of the argument.
11.2
The top matter
275
The typeset \translator is placed on the last page of the typeset article, before the address(es). There can be more than one translator. Each should be given as the argument of a separate \translator command. Example: \translator{Harry~M. Goldstein}
Rule
Dedication Command: \dedicatory Separate lines with \\ The typeset dedication is placed under the author(s). Example: \dedicatory{To the memory of my esteemed friend and teacher,\\ Harry~M. Goldstein}
Rule
Date Command: \date The typeset \date is placed on the front page of the typeset article as a footnote. Examples: \date{January 22, 2006} You can use the \today command to get today’s date: \date{\today} Do not use this when you submit an article; specify the submission date. To suppress the date, use \date{} or omit the \date command entirely.
11.2.2 Author information There are seven pieces of information about yourself.
276
Rule
Chapter 11
The AMS article document class
Author Command: \author Optional argument: Short form of the name for the running head The typeset author is placed on the front page of the typeset article. Examples: An author: \author{George~A. Menuhin} An author with a short form of the name for the running head: \author[G.\,A. Menuhin]{George~A. Menuhin} Section 11.2.4 discusses how to specify multiple authors.
Rule
Contributor Command: \contrib Optional argument: Describing the contribution The typeset contributor’s name is placed on the front page of the typeset article. This command is very recent. It was introduced in amsart version 2.20. Examples: A contributor authoring an appendix: \contrib[with an appendix by]{John Blaise} If this appendix has two authors: \contrib[with an appendix by]{J. Blaise} \contrib[]{W. Blaise} This typesets (with author G. A. Menuhin) the author line as G. A. MENUHIN, WITH AN APPENDIX BY J. BLAISE AND W. BLAISE
Contributors can have addresses, current addresses, etc., just like authors.
11.2
Rule
The top matter
277
Address Command: \address Separate lines with \\ Optional argument: Name of author The typeset address is placed at the end of the typeset article. Example: D EPARTMENT OF A PPLIED M ATHEMATICS , U NIVERSITY OF W INNEBAGO , W INNEBAGO , MN 53714
which is typed as \address{Department of Applied Mathematics\\ University of Winnebago\\ Winnebago, MN 53714} Notice that LATEX replaces the \\ line separators with commas. If there are several authors, you can use the author’s name as an optional argument of \address to avoid ambiguity. See Example 4 in Section 11.2.5 (page 283) for a complete example.
Rule
Current address Command: \curraddr Separate lines with \\ Optional argument: name of author The typeset current address is placed at the end of the typeset article. Example: Current address: Department of Mathematics, University of York, Heslington, York, England
is typed as \curraddr{Department of Mathematics\\ University of York\\ Heslington, York, England}
278
Chapter 11
The AMS article document class
If there are several authors, you can use the author’s name as an optional argument of \curraddr to avoid ambiguity; for some examples, see Section 11.2.5.
Rule
Email address Command: \email Optional argument: Name of author The typeset email address is placed at the end of the typeset article. Example: \email{
[email protected]}
Tip
Some email addresses contain the special underscore character ( _ ). Recall (see Section 5.4.4) that you have to type \_ to get _. Example: \email{George\
[email protected]}
Tip
Some older email addresses contain the percent symbol (%); recall that you have to type \% to get % (see Section 5.4.4). Example: \email{h1175moy\%
[email protected]}
Rule
Web (home) page (URL) Command: \urladdr Optional argument: Name of author
11.2
The top matter
279
The typeset Web (home) page is placed at the end of the typeset article. Example: \urladdr{http://www.maths.umanitoba.ca/homepages/gratzer/}
Tip
Many Internet addresses contain the tilde (~), indicating the home directory of the user. Type ~ to get ~ and not \~, as recommended in Section 5.4.4. $\sim$ is also unacceptable. Example: \urladdr{http://kahuna.math.hawaii.edu/~ralph/}
Rule
Research support or other acknowledgments Command: \thanks Do not specify linebreaks. Terminate the sentence with a period. The typeset research support or other acknowledgments is placed on the front page of the typeset article as an unmarked footnote. Example: \thanks{Supported in part by NSF grant PAL902466.} A \thanks{} command is ignored in typesetting.
11.2.3 AMS information The AMS requires that you supply two more pieces of information about the article. The following are collected at the bottom of the first page as unmarked footnotes along with the arguments of the \thanks and \date commands.
Rule
AMS subject classifications Command: \subjclass Optional argument: 2000—the default is 1991.
280
Chapter 11
The AMS article document class
amsart supplies the phrase 1991 Mathematics Subject Classification and a period at the end of the subject classification—with the optional argument 2000, the phrase is 2000 Mathematics Subject Classification The argument should be either a fivecharacter code or the phrase Primary: followed by a fivecharacter code, a semicolon, the phrase Secondary: and one or more additional fivecharacter codes. The typeset AMS subject classifications is placed at the bottom of the front page of the typeset article as a footnote. Examples: \subjclass[2000]{06B10} \subjclass[2000]{Primary: 06B10; Secondary: 06D05} The current subject classification scheme for mathematics was adopted in 2000, making the 1991 classification scheme obsolete. Thus, 2000 should be considered as a compulsory optional argument—maybe the only one in all of LATEX. The current subject classification scheme, MSC 2000, is available from the AMS Web site http://www.ams.org/ Search for MSC. Or in the Author Resource Center click on MSC.
Rule
Keywords Command: \keywords Do not indicate line breaks. amsart supplies the phrase Key words and phrases. and a period at the end of the list of keywords. The typeset keywords are placed on the front page of the typeset article as a footnote. Example: \keywords{Complete lattice, distributive lattice, complete congruence, congruence lattice} Keywords are optional for many journals.
11.2
The top matter
281
Further footnotes An additional \thanks command creates an unmarked footnote. Examples: \thanks{This is a preliminary version of this article, prepared for the Second Annual Meeting of the Statistical Association of Winnebago.} \thanks{This article is in final form, and no version of it will be submitted elsewhere.}
11.2.4 Multiple authors If an article has several authors, repeat the author information commands for each one. Take care that the email address follows the address. If two authors share the same address, omit the \address command for the second author, who can still have a different email address and Web home page. An additional \thanks command for the first author should precede any \thanks commands for the second author. Since the footnotes are not marked, the argument of the \thanks command for research support should contain a reference to the author: \thanks{The research of the first author was supported in part by NSF grant PAL902466.} \thanks{The research of the second author was supported by the Hungarian National Foundation for Scientific Research, under Grant No.~9901.} Finally, if an article has more than two authors, supply the author information for each author as usual, but explicitly specify the running heads with the \markleft command: \markleft{first author ET AL.} where first author must be all capitals. If there are multiple authors, sometimes it may not be clear whose address, current address, email address, or Web home page is being given. In such cases you can give the name of the author as an optional argument for these commands. For example, Email address, Ernest T. Moynahan:
[email protected] is typed as \email[Ernest~T. Moynahan]{
[email protected]} See also Example 4 in Section 11.2.5.
282
11.2.5
Chapter 11
The AMS article document class
Examples The following examples show typical top matter commands and can be found in the topmat.tpl file in the samples folder (see page 4). Example 1 One author. % Article information \title[Completesimple distributive lattices] {A construction of completesimple\\ distributive lattices} \date{\today} % Author information \author{George~A. Menuhin} \address{Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \email{
[email protected]} \urladdr{http://math.uwinnebago.edu/homepages/menuhin/} \thanks{This research was supported by the NSF under grant number 23466.} % AMS information \keywords{Complete lattice, distributive lattice, complete congruence, congruence lattice} \subjclass[2000]{Primary: 06B10; Secondary: 06D05} In the \title command, supplying the optional argument for the running head is the rule, not the exception. The only required item is \title. If it is missing, you get the strange error message: ! Undefined control sequence. <argument> \shorttitle l.49 \maketitle Example 2 Two authors but only the first has a Web home page. I only show the author information section here. The other commands are the same as in Example 1. % Author information \author{George~A. Menuhin} \address{Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714}
11.2
The top matter
283
\email{
[email protected]} \urladdr{http://math.uwinnebago.edu/homepages/menuhin/} \thanks{The research of the first author was supported by the NSF under grant number 23466.} \author{Ernest~T. Moynahan} \address{Mathematical Research Institute of the Hungarian Academy of Sciences\\ Budapest, P.O.B. 127, H1364\\ Hungary} \email{h1175moy\%
[email protected]} \thanks{The research of the second author was supported by the Hungarian National Foundation for Scientific Research, under Grant No. 9901.} Example 3 Two authors, same department. I only show the author information section here. The other commands are identical to those in Example 1. % Author information \author{George~A. Menuhin} \address{Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \email[George~A. Menuhin]{
[email protected]} \urladdr[George~A. Menuhin]% {http://math.uwinnebago.edu/homepages/menuhin/} \thanks{The research of the first author was supported by the NSF under grant number~23466.} \author{Ernest~T. Moynahan} \email[Ernest~T. Moynahan]{
[email protected]} \thanks{The research of the second author was supported by the Hungarian National Foundation for Scientific Research, under Grant No. 9901.} Note that the second author has no \address. Example 4 Three authors, the first two from the same department, the second and third with email addresses and research support. I only show the author information section. The other commands are unchanged. There are various ways of handling this situation. This example shows one solution. % Author information \author{George~A. Menuhin}
284
Chapter 11
The AMS article document class
\address[George~A. Menuhin and Ernest~T. Moynahan] {Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \email[George~A. Menuhin]{
[email protected]} \urladdr[George~A. Menuhin]% {http://math.uwinnebago.edu/homepages/menuhin/} \thanks{The research of the first author was supported by the NSF under grant number 23466.} \author{Ernest~T. Moynahan} \email[Ernest~T. Moynahan]{
[email protected]} \thanks{The research of the second author was supported by the Hungarian National Foundation for Scientific Research, under Grant No. 9901.} \author{Ferenc~R. Richardson} \address[Ferenc~R. Richardson] {Department of Mathematics\\ California United Colleges\\ Frasco, CA 23714} \email[Ferenc~R. Richardson]{
[email protected]} \thanks{The research of the third author was supported by the NSF under grant number 23466.}
Tip
The most common mistake in the top matter is the misspelling of a command name; for instance, \adress. LATEX sends the error message ! Undefined control sequence. l.37 \adress {Computer Science Department\\ which tells you exactly what you mistyped. Similarly, if you drop a closing brace, as in \email{
[email protected] you are told clearly what went wrong. Because the top matter commands are short (see Section 5.3.3), LATEX gives the error message Runaway argument? {
[email protected] \thanks {The research of th\ETC. !File ended while scanning use of \\email. If you drop an opening brace,
11.3
The sample article
285
\author George~A. Menuhin} you get the error message ! Too many }’s. l.43 \author George~A. Menuhin} If you enclose an optional argument in braces instead of brackets, \title{Completesimple distributive lattices}% {A construction of completesimple\\ distributive lattices} LATEX uses the short title as the title and the real title is typeset before the title of the typeset article.
11.2.6 Abstract As we discussed in Section 10.3.1, you type the abstract in an abstract environment, which you place as the last item before the \maketitle command. The abstract should be selfcontained, so do not include crossreferences and do not cite from the bibliography. Avoid userdefined commands. If you place the abstract after the \maketitle command, LATEX typesets it wherever it happens to be and sends a warning.
11.3
The sample article sampart.tex is the source file for our more advanced sample article using amsart (in the samples folder, see page 4). A simpler article using amsart is presented in Part I (see Section 4.1). The typeset sampart.tex is shown on the following three pages.
286
Chapter 11
The AMS article document class
A CONSTRUCTION OF COMPLETESIMPLE DISTRIBUTIVE LATTICES GEORGE A. MENUHIN Abstract. In this note we prove that there exist completesimple distributive lattices, that is, complete distributive lattices in which there are only two complete congruences.
1. Introduction In this note we prove the following result: Main Theorem. There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations. 2. The Dh2i construction For the basic notation in lattice theory and universal algebra, see Ferenc R. Richardson [5] and George A. Menuhin [2]. We start with some definitions: Definition 1. Let V be a complete lattice, and let p = [u, v] be an interval of V . Then p is called completeprime if the following three conditions are satisfied: (1) u is meetirreducible but u is not completely meetirreducible; (2) v is joinirreducible but v is not completely joinirreducible; (3) [u, v] is a completesimple lattice. Now we prove the following result: Lemma 1. Let D be a complete distributive lattice satisfying conditions (1) and (2). Then Dh2i is a sublattice of D2 ; hence Dh2i is a lattice, and Dh2i is a complete distributive lattice satisfying conditions (1) and (2). Proof. By conditions (1) and (2), Dh2i is a sublattice of D2 . Hence, Dh2i is a lattice. Since Dh2i is a sublattice of a distributive lattice, Dh2i is a distributive lattice. Using the characterization of standard ideals in Ernest T. Moynahan [3], Dh2i has a zero and a unit element, namely,Wh0, 0i and h1, 1i. To show that Dh2iWis complete, h2i let ∅ 6= A ⊆ Dh2i , and let a = A in D2 . If a ∈ Dh2i , then a = W A in D ; otherwise, a is of the form hb, 1i for some V b ∈ D with b < 1. Now A = h1, 1i in D2 and the dual argument shows that A also exists in D2 . Hence D is complete. Conditions (1) and (2) are obvious for Dh2i . Corollary 1. If D is completeprime, then so is Dh2i . Date: March 15, 2006. 2000 Mathematics Subject Classification. Primary: 06B10; Secondary: 06D05. Key words and phrases. Complete lattice, distributive lattice, complete congruence, congruence lattice. Research supported by the NSF under grant number 23466. 1
11.3
287
The sample article 2
GEORGE A. MENUHIN
The motivation for the following result comes from SooKey Foo [1]. Lemma 2. Let Θ be a complete congruence relation of Dh2i such that (2.1)
h1, di ≡ h1, 1i (mod Θ),
for some d ∈ D with d < 1. Then Θ = ι. Proof. Let Θ be a complete congruence relation of Dh2i satisfying (2.1). Then Θ = ι. 3. The Π∗ construction The following construction is crucial to our proof of the Main Theorem: Definition 2. Let Di , for i ∈ I, be complete distributive lattices satisfying condition (2). Their Π∗ product is defined as follows: Π∗ (Di  i ∈ I) = Π(Di−  i ∈ I) + 1; that is, Π (Di  i ∈ I) is Π(Di−  i ∈ I) with a new unit element. ∗
Notation. If i ∈ I and d ∈ Di− , then i
h. . . , 0, . . . , d, . . . , 0, . . . i is the element of Π∗ (Di  i ∈ I) whose ith component is d and all the other components are 0. See also Ernest T. Moynahan [4]. Next we verify: Theorem 1. Let Di , for i ∈ I, be complete distributive lattices satisfying condition (2). Let Θ be a complete congruence relation on Π∗ (Di  i ∈ I). If there exist i ∈ I and d ∈ Di with d < 1i such that for all d ≤ c < 1i , (3.1)
i
i
i
i
h. . . , 0, . . . , d, . . . , 0, . . . i ≡ h. . . , 0, . . . , c, . . . , 0, . . . i (mod Θ),
then Θ = ι. Proof. Since (3.2)
h. . . , 0, . . . , d, . . . , 0, . . . i ≡ h. . . , 0, . . . , c, . . . , 0, . . . i (mod Θ),
and Θ is a complete congruence relation, it follows from condition (3) that i
(3.3)
h. . . , d, . . . , 0, . . . i _ i ≡ (h. . . , 0, . . . , c, . . . , 0, . . . i  d ≤ c < 1) ≡ 1 (mod Θ).
Let j ∈ I for j 6= i, and let a ∈ Dj− . Meeting both sides of the congruence (3.2) j
with h. . . , 0, . . . , a, . . . , 0, . . . i, we obtain i
(3.4)
j
0 = h. . . , 0, . . . , d, . . . , 0, . . . i ∧ h. . . , 0, . . . , a, . . . , 0, . . . i j
≡ h. . . , 0, . . . , a, . . . , 0, . . . i (mod Θ).
Using the completeness of Θ and (3.4), we get: _ j 0 ≡ (h. . . , 0, . . . , a, . . . , 0, . . . i  a ∈ Dj− ) = 1 (mod Θ), hence Θ = ι.
288
Chapter 11
The AMS article document class
COMPLETESIMPLE DISTRIBUTIVE LATTICES
3
Theorem 2. Let Di for i ∈ I be complete distributive lattices satisfying conditions (2) and (3). Then Π∗ (Di  i ∈ I) also satisfies conditions (2) and (3). Proof. Let Θ be a complete congruence on Π∗ (Di  i ∈ I). Let i ∈ I. Define i
b i = {h. . . , 0, . . . , d, . . . , 0, . . . i  d ∈ D− } ∪ {1}. D i b i is a complete sublattice of Π∗ (Di  i ∈ I), and D b i is isomorphic to Di . Then D b Let Θi be the restriction of Θ to Di . b i , and hence Θi is ω or ι. If Θi = ρ for all Since Di is completesimple, so is D i ∈ I, then Θ = ω. If there is an i ∈ I, such that Θi = ι, then 0 ≡ 1 (mod Θ), hence Θ = ι. The Main Theorem follows easily from Theorems 1 and 2. References [1] SooKey Foo, Lattice Constructions, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. [2] George A. Menuhin, Universal algebra. D. Van Nostrand, Princeton, 1968. [3] Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. [4] , Ideals and congruence relations in lattices. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl. 9 (1957), 417–434 (Hungarian). [5] Ferenc R. Richardson, General lattice theory. Mir, Moscow, expanded and revised ed., 1982 (Russian). Computer Science Department, University of Winnebago, Winnebago, MN 53714 Email address:
[email protected] URL: http://math.uwinnebago.edu/homepages/menuhin/
11.3
The sample article
289
The next two pages show the first page of the source file facing a fragment of the first page of the typeset article. It displays the part of this first typeset page that comes from the facing source file. On the two pages after the facing first pages, some parts of the source file and the typeset version are shown juxtaposed, so that you can see how the markedup source file becomes the typeset article.
290
Chapter 11
The AMS article document class
% Sample file: sampart.tex % The sample article for the amsart document class \documentclass{amsart} \usepackage{amssymb,latexsym} \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{corollary}{Corollary} \newtheorem*{main}{Main~Theorem} \newtheorem{lemma}{Lemma} \newtheorem{proposition}{Proposition} \theoremstyle{definition} \newtheorem{definition}{Definition} \theoremstyle{remark} \newtheorem*{notation}{Notation} \numberwithin{equation}{section} \begin{document} \title[Completesimple distributive lattices] {A construction of completesimple\\ distributive lattices} \author{George~A. Menuhin} \address{Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \email{
[email protected]} \urladdr{http://math.uwinnebago.edu/homepages/menuhin/} \thanks{Research supported by the NSF under grant number 23466.} \keywords{Complete lattice, distributive lattice, complete congruence, congruence lattice} \subjclass[2000]{Primary: 06B10; Secondary: 06D05} \date{March 15, 2006} \begin{abstract} In this note we prove that there exist \emph{completesimple distributive lattices,} that is, complete distributive lattices in which there are only two complete congruences. \end{abstract} \maketitle \section{Introduction}\label{S:intro} In this note we prove the following result: \begin{main} There exists an infinite complete distributive lattice~$K$ with only the two trivial complete congruence relations. \end{main} \section{The $D^{\langle 2 \rangle}$ construction}\label{S:Ds} For the basic notation in lattice theory and universal algebra, see Ferenc~R. Richardson~\cite{fR82} and George~A. Menuhin~\cite{gM68}. We start with some definitions:
11.3
291
The sample article
A CONSTRUCTION OF COMPLETESIMPLE DISTRIBUTIVE LATTICES GEORGE A. MENUHIN Abstract. In this note we prove that there exist completesimple distributive lattices, that is, complete distributive lattices in which there are only two complete congruences.
1. Introduction In this note we prove the following result: Main Theorem. There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations. 2. The Dh2i construction For the basic notation in lattice theory and universal algebra, see Ferenc R. Richardson [5] and George A. Menuhin [2]. We start with some definitions:
Date: March 15, 2006. 2000 Mathematics Subject Classification. Primary: 06B10; Secondary: 06D05. Key words and phrases. Complete lattice, distributive lattice, complete congruence, congruence lattice. Research supported by the NSF under grant number 23466. 1
292
Chapter 11
The AMS article document class
Since $D^{\langle 2 \rangle}$ is a sublattice of a distributive lattice, $D^{\langle 2 \rangle}$ is a distributive lattice. Using the characterization of standard ideals in Ernest~T. Moynahan~\cite{eM57}, $D^{\langle 2 \rangle}$ has a zero and a unit element, namely, $\langle 0, 0 \rangle$ and $\langle 1, 1 \rangle$. To show that $D^{\langle 2 \rangle}$ is complete, let $\varnothing \ne A \subseteq D^{\langle 2 \rangle}$, and let $a = \bigvee A$ in $D^{2}$. If $a \in D^{\langle 2 \rangle}$, then $a = \bigvee A$ in $D^{\langle 2 \rangle}$; otherwise, $a$ is of the form $\langle b, 1 \rangle$ for some $b \in D$ with $b < 1$. Now $\bigvee A = \langle 1, 1\rangle$ in $D^{2}$ and the dual argument shows that $\bigwedge A$ also exists in $D^{2}$. Hence $D$ is complete. Conditions \eqref{mi} and~\eqref{ji} are obvious for $D^{\langle 2 \rangle}$. \end{proof} \begin{corollary}\label{C:prime} If $D$ is completeprime, then so is $D^{\langle 2 \rangle}$. \end{corollary} The motivation for the following result comes from SooKey Foo~\cite{sF90}. \begin{lemma}\label{L:ccr} Let $\Theta$ be a complete congruence relation of $D^{\langle 2 \rangle}$ such that \begin{equation}\label{E:rigid} \langle 1, d \rangle \equiv \langle 1, 1 \rangle \pmod{\Theta}, \end{equation} for some $d \in D$ with $d < 1$. Then $\Theta = \iota$. \end{lemma}
11.3
293
The sample article
b i , and hence Θi is ω or ι. If Θi = ρ for all Since Di is completesimple, so is D i ∈ I, then Θ = ω. If there is an i ∈ I, such that Θi = ι, then 0 ≡ 1 (mod Θ), hence Θ = ι. The Main Theorem follows easily from Theorems 1 and 2. References [1] SooKey Foo, Lattice Constructions, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. [2] George A. Menuhin, Universal algebra . D. Van Nostrand, Princeton, 1968. [3] Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. [4] , Ideals and congruence relations in lattices. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl. 9 (1957), 417–434 (Hungarian). [5] Ferenc R. Richardson, General lattice theory . Mir, Moscow, expanded and revised ed., 1982 (Russian). Computer Science Department, University of Winnebago, Winnebago, MN 53714 Email address:
[email protected] URL: http://math.uwinnebago.edu/homepages/menuhin/
Since $D_{i}\) is completesimple, so is $\widehat{D}_{i}$, and hence $\Theta_{i}$ is $\omega$ or $\iota$. If $\Theta_{i} = \rho$ for all $i \in I$, then $\Theta = \omega$. If there is an $i \in I$, such that $\Theta_{i} = \iota$, then $0 \equiv 1 \pmod{\Theta}$, hence $\Theta = \iota$. \end{proof} The Main Theorem follows easily from Theorems \ref{T:P*} and~\ref{T:P*a}. \begin{thebibliography}{9} \bibitem{sF90} SooKey Foo, \emph{Lattice Constructions}, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. \bibitem{gM68} George~A. Menuhin, \emph{Universal algebra},. D.~Van Nostrand, Princeton, 1968. \bibitem{eM57} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \textbf{8} (1957), 455460. \bibitem{eM57a} \bysame, \emph{Ideals and congruence relations in lattices}. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K\"{o}zl. \textbf{9} (1957), 417434 (Hungarian). \bibitem{fR82} Ferenc~R. Richardson, \emph{General lattice theory},. Mir, Moscow, expanded and revised ed., 1982 (Russian). \end{thebibliography} \end{document}
294
11.4
Chapter 11
The AMS article document class
Article templates In this section, we create a template you can use for your amsart articles. A template is a readonly file. Open it with a text editor and save it under a different name. You can then start to write your new article using the new file, without having to remember the details governing the preamble and the top matter. Create the template, which contains a customized preamble and top matter with sample bibliographic items, in several steps. Step 1 In your text editor, open the amsart.tpl document from the samples folder (see page 4) and save it in your work subfolder as myams.tpl. Alternatively, type in the lines as shown in this section. The first few lines of the file are % Sample file: amsart.tpl % Preamble \documentclass{amsart} \usepackage{amssymb,latexsym} Notice the use of commented out lines (lines that start with %) that have been added as comments about the file. Edit line 1 to read % Template file: myams.tpl The lines \documentclass{amsart} \usepackage{amssymb,latexsym} specify the amsart document class and the use of the amssymb and latexsym packages to gain access by name to all the symbols listed in Appendix B. Step 2 After the \usepackage command, there are sets of proclamation definitions corresponding to the examples in Section 6.4.2. Choose Option 5 for myams.tpl by deleting all the lines related to the other options. You are left with the lines % % % % % %
Theorems, corollaries, lemmas, and propositions, in the most emphatic (plain) style. All are numbered separately. There is a Main Theorem in the most emphatic (plain) style, unnumbered. There are definitions, in the less emphatic(definition) style. There are notations, in the least emphatic (remark) style, unnumbered.
11.4
Article templates
295
\theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{corollary}{Corollary} \newtheorem*{main}{Main Theorem} \newtheorem{lemma}{Lemma} \newtheorem{proposition}{Proposition} \theoremstyle{definition} \newtheorem{definition}{Definition} \theoremstyle{remark} \newtheorem*{notation}{Notation} Step 3 Two more choices are presented. You can have either one or two authors—for more complex situations, see Section 11.2.4. For the myams.tpl template, choose one author by deleting everything between %
Two authors
and %
End Two authors
You are left with \begin{document} % One author \title[shorttitle]{titleline1\\ titleline2} \author{name} \address{line1\\ line2\\ line3} \email{
[email protected]} \urladdr{http://homepage} \thanks{thanks} % End one author \keywords{keywords} \subjclass[2000]{Primary: subject; Secondary: subject} \date{date} \begin{abstract} abstract
296
Chapter 11
The AMS article document class
\end{abstract} \maketitle \begin{thebibliography}{99} \end{thebibliography} \end{document} In the top matter, fill in your own personal information. For instance, I edited \author{name} to read \author{George~Gr\"{a}tzer} I also edited \address, \email, \urladdr, and \thanks. After editing, I had the following: % top matter \title[shorttitle]{titleline1\\ titleline2} \author{George~Gr\"{a}tzer} \address{University of Manitoba\\ Department of Mathematics\\ Winnipeg, MB R3T 2N2\\ Canada} \email{
[email protected]} \urladdr{http://server.maths.umanitoba.ca/homepages/gratzer/} \thanks{Research supported by the NSERC of Canada.} \keywords{keywords} \subjclass[2000]{Primary: subject; Secondary: subject} \date{date} \begin{abstract} abstract \end{abstract} \maketitle \begin{thebibliography}{99} \end{thebibliography} \end{document}
11.5
297
Options
Since this template is meant to be used for all my future articles, I do not edit the lines that change from article to article (\title, \keywords, and so on). Remember that the short title is for running heads, the title shown at the top of every oddnumbered page other than the title page. If the title of your article is only one line long, delete the separation mark \\ and the second line, except for the closing brace. If the full title of your article is short, delete [shorttitle]. Now save myams.tpl. I saved my template under the name ggamsart.tpl (in the samples folder, see page 4). You can also make an additional template with two authors to be used as a template for joint articles. Note that at the end of the template, just before the line \end{document}, there are two lines: \begin{thebibliography}{99} \end{thebibliography} The argument of \begin{thebibliography} should be 9 if there are fewer than 10 references, 99 with 10–99 references, and so forth. We discuss how to format bibliographic items in Sections 4.2.4 and 10.5.1. The templates for bibliographic items are listed after the \end{document} line. To make sure that you do not overwrite your template, I recommend that you make it readonly. How you do this depends on your computer’s operating system. You should modify the template you create in this section to the template of the journal you submit your article to. In the samples folder, you find the AMS template for the Proceedings of the AMS, called amsproc.template.
11.5
Options The amsart document class supports a number of options, affecting many attributes. For each attribute there is a default value that is used if a value is not specified. Font size Options:
9pt 10pt 11pt 12pt
default
This option declares the default font size. You may want to use the 12pt option for proofreading: \documentclass[12pt]{amsart} Remember, however, that changing the font size changes the line breaks, so changing the 12pt option back to 10pt may require that you make some adjustments in the text (see Section 2.3).
298
Chapter 11
The AMS article document class
Paper size Options:
letterpaper legalpaper a4paper
(8.5 inches by 11 inches) (8.5 inches by 14 inches) (210 mm by 297 mm)
default
Equations and equation numbers A number of options deal with the placement of equations and equation numbers. Options:
leqno reqno
default
By default, equation numbers are placed on the left, the default leqno option. The reqno option places the equation numbers on the right. Option:
fleqn
This option positions equations a fixed distance from the left margin rather than centering them. The fleqn option is typically used in conjunction with the reqno option. Here is how an equation looks with the fleqn and reqno options: Z π sin x dx = 2 (1) 0
typed as \begin{equation}\label{E:firstInt} \int_{0}^{\pi} \sin x \, dx = 2 \end{equation} Options:
tbtags centertags
default
The tbtags option uses toporbottom tags for a split environment, that is, it places the equation number level with the last line if numbers are on the right, or level with the first line if the numbers are on the left:
(1)
f = (x1 x2 x3 x4 x5 x6 )2 = (x1 x2 x3 x4 x5 + x1 x3 x4 x5 x6 + x1 x2 x4 x5 x6 + x1 x2 x3 x5 x6 )2 = (x1 x2 x3 x4 + x1 x2 x3 x5 + x1 x2 x4 x5 + x1 x3 x4 x5 )2
The centertags option (the default) vertically centers the equation number in a split subsidiary math environment.
11.5
299
Options Limits Options:
intlimits nointlimits
default
The intlimits option places the subscripts and superscripts of integral symbols above and below the integral symbol rather than on the side in a displayed math formula— with this option you can use the \nolimit command to disable the option for one integral. The nointlimits option positions the subscripts and superscripts of integral symbols on the side. Options:
sumlimits nosumlimits
default
The sumlimits option places the subscripts and superscripts of large operators, such P Q ` N L as , , , , , above and below the large operator in a displayed math formula. nosumlimits positions them on the side (see Table 7.5 and Section B.7.1). Options:
namelimits nonamelimits
default
The namelimits option places the subscripts and superscripts of operators with limits such as det, inf, lim, max, min, and so on, above and below the operator in a displayed math formula. nonamelimits positions them on the side (see Tables 7.2, 7.3, and Section B.7). Twosided printing Options:
twoside oneside
default
The twoside option formats the output for printing on both sides of the paper. The alternative is the oneside option. This option influences running heads, the placement of page numbers, and so on. Twocolumn printing Options:
twocolumn onecolumn
default
The twocolumn option typesets the document in two columns. Title page Options:
titlepage notitlepage
default
The titlepage option creates a separate title page including the abstract. The notitlepage option splits the top matter between the first and last pages of the typeset article.
300
Chapter 11
The AMS article document class
Draft Options:
draft final
default
The draft option prints a slug in the margin next to each line that is too wide. The final option does not. Note that this option is passed on to some packages, such as graphicx. Fonts Option:
noamsfonts
With this option, the document class does not load the packages necessary for the use of the AMSFonts font set. Option:
psamsfonts
The psamsfonts option tells LATEX to use the PostScript version of the AMSFonts set. No math Option:
nomath
By default, amsart loads the amsmath package (which, in turn, loads three more math packages). If you want to use the title page and related features without the math features, you can use the nomath option.
11.6
The AMS packages If you follow the recommendation of this book and begin each article with \documentclass{amsart} \usepackage{amssymb,latexsym} then you can safely ignore most of the information in this section. There are two minor exceptions, the packages amsxtra and upref. However, if you use a document class that does not load the same packages that amsart loads, then you have to load the packages needed for your work. Typically, you have \usepackage{amsmath,amsfonts,amsthm} \usepackage{amssymb,latexsym} as a minimum. The AMS distribution contains many packages that can be loaded together or by themselves.
11.6
The AMS packages
301
Math enhancements amsmath The primary math enhancement package, which loads the four packages, amsgen, amsbsy, amsopn, and amstext. amsbsy Provides two commands for the use of bold math symbols, \boldsymbol and \pmb (see Section 8.3.3). amscd Commands for creating simple commutative diagrams (see Section 9.8). amsgen An auxiliary package that is never invoked directly. It is loaded by all the AMS math packages (except for upref). amsopn Provides operator names and also the \DeclareMathOperator command for defining new ones (see Section 7.6). amstext Defines the \text command and redefines commands such as \textrm and \textbf to behave like the \text command (see Section 7.4.6). amsxtra Provides the “sp” math accents (see Sections 7.7 and B.8) and loads the amsmath package. upref Ensures that the \ref command always produces upright numbers. AMSFonts amsfonts Contains the basic commands needed to utilize the AMSFonts. It also defines the \mathfrak command which makes the Euler Fraktur math alphabet available (see Section 8.3.2). If you use the PostScript AMSFonts font set, you should load this package with the option \usepackage[psamsfonts]{amsfonts} In addition, if you want to use the 12pt document class option, then you must also load the exscale package (see Section 12.3): \usepackage{exscale} amssymb Defines the symbol names for amsfonts. It loads amsfonts. eucal Replaces the calligraphic math alphabet with the Euler Script math alphabet (see Section 8.3.2). If you load it with the option mathscr, as in \usepackage[mathscr]{eucal} then both the \mathscr and the \mathcal commands are available, so you can have both CE and CE, typed as
302
Chapter 11
The AMS article document class
$\mathcal{C}\mathcal{E}$ and $\mathscr{C}\mathscr{E}$ eufrak Defines the Euler Fraktur math alphabet (see Section 8.3.2). Loading packages amsart contains code to provide more flexible formatting of proclamations and the proof environment (see Sections 6.4.2 and 6.5). By loading the amsthm package you can add this functionality to a nonAMS document class. The amsthm package loads the amsgen package. amsart loads four packages from the math enhancements group, the amsmath, amsbsy, amstext, amsopn, and amsgen packages, and the amsfonts package from the AMSFonts group. A typical article using the legacy article document class (see Section 12.1) and the AMS enhancements would normally have \documentclass{article} \usepackage{amsmath}% math enhancements \usepackage{amssymb,latexsym}% AMSFonts and LaTeX symbol names \usepackage{amsthm}% proclamations with style and perhaps the following: \usepackage{eucal}% Euler Script Note that it is not critical for you to remember which packages load others. No harm is done if you type \usepackage{amsmath} \usepackage{amsbsy} The amsbsy package is loaded by the amsmath package, and the \usepackage{amsbsy} line is ignored by LATEX. All the math related options of amsart (see Section 11.5) are also options of the amsmath package. So, for instance, if you want the equation numbers on the right, load amsmath with the reqno option: \usepackage[reqno]{amsmath} Multiple indices The AMS distribution also contains the package amsmidx for creating multiple indices. This package is discussed in Section 17.5.
CHAPTER
12 Legacy document classes
Even though the AMS spent a few decades refining the amsart document class, some of the legacy document classes of LATEX are still around. If you want to whip up a quick report or write up a research note, the legacy article or report document classes may serve you well. In this chapter, we discuss some of the legacy LATEX document classes. We take up the book document class in Chapter 18. We do not discuss the slides document class for preparing slides. It is now considered obsolete. Use instead the beamer class which we discuss in Chapter 14 or the FoilTEX class discussed in Section 4.4. We conclude this chapter with a description of the components of the standard LATEX distribution.
12.1
Articles and reports The article and report document classes are very similar. There are two substantive differences to remember: 1. The report document class provides a separate page for the abstract by default, the article document class does not.
304
Chapter 12
Legacy document classes
2. The report document class has two additional sectioning commands, \chapter and \part. We discuss these commands in Section 18.1.1. In the samples folder (see page 4) is the document legacyarticle.tex, a variant of the introductory sample article, intrart.tex of Chapter 4. The first page of this is typeset on the following page.
12.1.1
Top matter For a detailed discussion of the top matter, refer to Chapter 10, in particular Sections 10.1 and 10.3—see also Figure 10.1. Here is the top matter of the legacy article: \title{A construction of completesimple\\ distributive lattices} \author{George~A. Menuhin\thanks{Research supported by the NSF under grant number 23466.}\\ Computer Science Department\\ Winnebago, MN 23714\\
[email protected]} \date{March 15, 2000} \maketitle There are four commands for the top matter: \title, \author, \thanks, and \date.
Rule
Top matter for the article document class 1. If necessary, break the title into separate lines with \\. Do not put a \\ at the end of the last line. 2. \thanks places a footnote at the bottom of the first page. If it is not needed, omit it. 3. Separate the lines of the address with \\. Do not put a \\ at the end of the last line. 4. Multiple authors are separated by \and. There is only one \author command, and it contains all the information—name, address, support—about all the authors. There is no \\ command before the \and command. 5. If there is no \date command, LATEX will insert the date on which you typeset the file (\date{\today} will produce the same result). If you do not want any date to appear, type \date{}. For a specific date, such as February 21, 2007, type \date{February 21, 2007}. 6. The \title command is the only required command. The others are optional.
12.1
305
Articles and reports
A construction of completesimple distributive lattices George A. Menuhin∗ Computer Science Department Winnebago, MN 23714
[email protected] March 15, 2006 Abstract In this note, we prove that there exist completesimple distributive lattices, that is, complete distributive lattices in which there are only two complete congruences.
1
Introduction
In this note, we prove the following result: Theorem 1 There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations.
2
The Π∗ construction
The following construction is crucial in the proof of our Theorem: Definition 1 Let Di , for i ∈ I, be complete distributive lattices satisfying condition (J). Their Π∗ product is defined as follows: Π∗ (Di  i ∈ I) = Π(Di−  i ∈ I) + 1; that is, Π∗ (Di  i ∈ I) is Π(Di−  i ∈ I) with a new unit element. Notation 1 If i ∈ I and d ∈ Di− , then h. . . , 0, . . . , d, . . . , 0, . . .i is the element of Π∗ (Di  i ∈ I) whose ith component is d and all the other components are 0. ∗ Research
supported by the NSF under grant number 23466.
1
306
Chapter 12
Legacy document classes
As you see, the rules for the \date command here differ slightly from the rules for the \date command in the amsart document class. However, the rules for the command \author here are very different from the rules for the \author command in the amsart document class. For two authors use the following template: \author{name1\thanks{support1}\\ address1line1\\ address1line2\\ address1line3 \and name2\thanks{support2}\\ address2line1\\ address2line2\\ address2line3} Note the use of the \and command, which separates the two authors. One more difference to keep in mind. Place the abstract after the \maketitle command.
12.1.2
Options The article and report document classes have a similar range of options. These are listed below. Font size 10pt default 11pt 12pt Each option declares the specified size to be the default font size. Options:
Paper size Options:
letterpaper legalpaper executivepaper a4paper a5paper b5paper
Draft Options:
draft final
default
(8.5 inches by 11 inches) (8.5 inches by 14 inches) (7.25 inches by 10.5 inches) (210 mm by 297 mm) (148 mm by 210 mm) (176 mm by 250 mm)
default
12.1
307
Articles and reports
The draft option places a slug in the margin next to each line that is too wide (see Section 2.3). The final option does not. Note that this option is passed on to some packages, such as graphicx. To prevent this, invoke graphicx with the final option. Landscape printing Option:
landscape
The landscape option typesets the document in landscape format, swapping the width and height of the paper. Twosided printing Options:
twoside oneside
default
The twoside option formats the output for printing on both sides of the paper. Twocolumn printing Options:
twocolumn onecolumn
default
The twocolumn option typesets the document in twocolumn format. This option has many problems. It is better to use the multicol package (see Section 12.3.1). Title page Options:
titlepage notitlepage
default for report default for article
The titlepage option creates a separate title page and places the abstract on a separate page. The notitlepage option places the title and the abstract together on the first page. Equations and equation numbers Options:
leqno reqno
default
The leqno option places any equation number in the document on the left side and reqno places them on the right. Option: fleqn The fleqn option sets displayed formulas flush left. This option is typically used in conjunction with the reqno option.
308
Chapter 12
Legacy document classes
[1] SooKey Foo. Lattice Constructions. PhD thesis, University of Winnebago, Winnebago, MN, December 1990. [2] George A. Menuhin. Universal Algebra. D. Van Nostrand, Princeton, 1968. [3] Ernest T. Moynahan. Ideals and congruence relations in lattices. II. Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl., 7:417–434, 1957. [4] Ernest T. Moynahan. Stone. On a problem of M.H. Acta Math. Acad. Sci. Hungar., 8:455–460, 1957. [5] Ferenc R. Richardson. General Lattice Theory. Mir, Moscow, expanded and revised edition, 1982. Figure 12.1: The openbib option. Bibliography Option:
openbib
The openbib option typesets the bibliography in a spread out “open” format (see Figure 12.1). Combinations Of course, these options can be combined with each other and are also used by most legacy document classes. For instance, \documentclass[12pt,a4paper,twoside,twocolumn]{report} produces a doublecolumned, twosided report on A4 paper, the European standard, at the 12point font size.
12.2
Letters The letter document class was developed for writing letters. One document can contain any number of letters, each in its own letter environment. In the following example (letter.tex in the samples folder) there is only a single letter: % Sample file: letter.tex \documentclass{letter}
12.2
Letters
309
\begin{document} \address{George Gr\"{a}tzer\\ Department of Mathematics\\ University of Manitoba\\ Winnipeg, MB, R3T 2N2\\ Canada} \signature{George Gr\"{a}tzer} \date{} \begin{letter}{Prof.~John Hurtig\\ Computer Science Department\\ University of Winnebago\\ Winnebago, Minnesota 23714} \opening{Dear John,} Enclosed you will find the first draft of the fiveyear plan. \closing{Friendly greetings,} \cc{Carla May\\ Barry Bold} \encl{Fiveyear plan} \ps{P.S. Remember our lunch meeting tomorrow! G.} \end{letter} \end{document} Figure 12.2 shows the typeset letter. The argument of the letter environment is the name and address of the recipient. It is a required argument and if it is omitted, you get an error message such as ! Incomplete \iffalse; all text was ignored after line 21. \fi l.21 \end{letter} As with all multiline arguments, the lines are separated by \\. The arguments of some commands may apply to all the letter environments in the document. Such commands should be placed before the first letter environment. In the example, \signature and \address are so placed. If the \date command is absent, today’s date is typeset. If you want no date, use an empty argument \date{}, as in the example. If you want all the letters in the same document to have the same date, the \date command should precede the first letter environment.
310
Chapter 12
Legacy document classes
George Gr¨atzer Department of Mathematics University of Manitoba Winnipeg, MB, R3T 2N2 Canada
Prof. John Hurtig Computer Science Department University of Winnebago Winnebago, Minnesota 23714 Dear John, Enclosed you will find the first draft of the fiveyear plan. Friendly greetings,
George Gr¨atzer cc: Carla May Barry Bold encl: Fiveyear plan P.S. Remember our lunch meeting tomorrow! G. Figure 12.2: A letter. Many of the options listed in Section 12.1.2 can also be invoked for the letter document class.
12.3
The LATEX distribution The LATEX distribution contains a number of document classes and packages, most of which you have probably received with your TEX software. If you find that you are missing some files, see Section E.1 on how to get them. The files of the LATEX distribution on CTAN are grouped in the directory /pub/tex/macros/latex into four subdirectories.
12.3
The LATEX distribution
311
base contains all the files necessary to install the system. As a rule, for every package, say, exscale, it contains two files, exscale.ins and exscale.dtx. Typesetting the first gives you exscale.sty and typesetting the second produces the user guide and the commented source code. Since most LATEX implementations install the content of the unpacked directory, this directory is not for the average user. doc contains LATEX documentation in PDF files and also the LATEX News. required contains the directories amslatex, babel, cyrillic, graphics, psnfss, tools. unpacked contains the unpacked LATEX distribution. Since most LATEX implementations install this, you may never need it. Of the packages and tex files included in the unpacked folder, the following should be of special interest to readers of this book. latexsym Some symbol definitions (see the tables in Appendix B). alltt The alltt environment, which is like the verbatim environment except that \, and {, } retain their usual meanings. exscale Scaled versions of the math extension font. makeidx Commands for producing indexes (see Chapter 17). showidx A package to allow you to typeset the index entries in the margin of your typeset document (see Section 17.1). nfssfont.tex Generates font tables for use with the \symbol command (see Section 5.4.4). There is also the file fixltx2e.sty in the unpacked directory (and the corresponding fixltx2e.dtx and fixltx2e.ins in the base directory). This file contains fixes to latex.ltx, the main LATEX file, and also some new commands that did not make it into the current release. The last fixltx2e.sty (Dec. 2005) contains two important additions. It complements the \textsuperscript command (see Section C.4) with a \textsubscript command. More importantly, it introduces a very useful new command \TextOrMath. This command has two arguments and it typesets the first in text and the second in math. For instance, if in newlattice.sty (see 15.3 ) you define \newcommand{\Gra}{\TextOrMath{$\alpha$\xspace}{\alpha}} then you can type \Gra in both test and math. Indeed. this is \Gra in text, and this is $\Gra  x^2$ in math
312
Chapter 12
Legacy document classes
typesets as this is α in text, and this is α − x2 in math If you want to use these commands, include the line \usepackage{fixltx2e} in the preamble. In the required folder there are some major software distributions related to LATEX. amslatex Discussed in detail in this book, this directory contains the AMS math packages and document classes, while the fontrelated AMS files are in the directory /texarchive/fonts/amsfonts/latex/ babel For typesetting languages other than American English. cyrillic For typesetting Cyrillic characters. graphicx For the inclusion and transformation of graphics and for typesetting in color (see Section 10.4.3). This package requires that you have a suitable printer driver. psnfss For typesetting with a wide range of PostScript fonts (see Section F.1). tools A range of tools for managing document production discussed in the next section. Each of these packages comes with its own documentation. They are also described in The LATEX Companion, 2nd edition [46].
12.3.1
Tools Some of these packages are so important that they could well have been incorporated into LATEX proper. Here is a brief listing. afterpage Implements the \afterpage command. The commands specified in its argument are expanded after the current page is output. array Contains extended versions of the array and tabular environments with many extra features. bm Gives access to bold math symbols. calc Allows algebraic manipulation of lengths and counter values when specifying lengths and counters.
12.3
The LATEX distribution
313
dcolumn Provides alignment on decimal points in tabular entries. It requires the array package. delarray Adds “large delimiters” around arrays. It requires the array package. enumerate Provides customized enumerate environments (see Sections 6.2.4 and also Sections 15.2.1, 15.3, 15.4). fileerr Helps with missing files. fontsmpl Produces a test file for displaying “font samples”. ftnright Places all footnotes in the righthand column of documents typeset with the twocolumn document class option. hhline Provides control over horizontal lines in tables. indentfirst Indents the first paragraph of each section. layout Shows the page layout defined by a document class (see Section 10.6). longtable Helps to create multipage tables. It does not require the array package, but it uses array’s extended features if both packages are loaded. multicol Provides multicolumn typesetting with some advanced features. rawfonts Preloads fonts using the old font names of LATEX 2.09. showkeys Selectively prints the labels used by \label, \ref, \cite, and so forth, in the margin (see Section 10.4.2). somedefs Elective handling of package options. It is used by the rawfonts package. tabularx Defines a variant of the tabular environment where all the columns are the same width. It requires the array package. theorem Allows the definition of proclamations in flexible formats. The AMS variant, the amsthm package, is discussed in Section 6.4.2. trace Tracing help for macro writers. varioref Provides smart as well as multilingual handling of page references. verbatim Extends the verbatim environment and provides the comment environment (see Sections 5.5.1 and 6.8). xr Creates crossreferences among documents. xspace Provides a “smart space” command that helps you avoid the common mistake of missing space after commands. It is mainly used in commands that expand to some text (see Section 15.1.1). All of these packages are discussed in The LATEX Companion, 2nd edition [46].
CHAPTER
13 PDF
13.1
documents
PostScript and PDF
13.1.1 PostScript PostScript is the preeminent platform and device independent pagedescription and programming language, introduced by Adobe Systems Inc. It describes the placement and shapes of all the elements in the document, including the fonts. Documents placed on the Web in PostScript format can be downloaded to any computer and print identically on all PostScript printers. Until the appearance of PDF, PostScript was the format of choice for sharing LATEX articles with diagrams or complex forms. There are a number of disadvantages to using PostScript files on the Web: The files tend to be very large. They cannot be viewed until the whole file has been downloaded. If a PostScript file does not include a particular font used in the document and you do not have that font installed on your computer, then another font—usually Courier— is substituted causing graphically unacceptable rendering.
318
13.1.2
Chapter 13 PDF documents PDF All of these concerns have been addressed by Adobe’s Portable Document Format (PDF). See Adobe Systems’ PDF Reference, Version 1.6, 5th edition [2] for a complete description of this file format. PDF is based on the PostScript language, with some important differences: PDF is much more concise than PostScript. A PDF file is normally about 10 percent of the size of the corresponding PostScript file.
Missing fonts are usually substituted by fonts with the same metrics, so that the size of the substituted text is the same as that of the original. In particular, there are no incorrect line breaks caused by the substitution. files allow partial inclusion of fonts. As a result, it is much easier to obtain permission to include proprietary fonts in PDF documents. PDF
PDF files can be downloaded and viewed in a Web browser one page at a time, without having to wait for the whole file to download first.
Many LATEX implementations, including the two discussed in Appendix A, produce PDF files. PDF
files and Adobe Acrobat Professional offer many nice features, including:
Efficient navigational tools Searching and indexing capability for documents and even for collections of documents Bookmarks Thumbnails of pages Limited editing Annotations (notes, text, and voice) and markups Hyperlinks to the same document or to another document or Web site (see Section 13.1.3) The inclusion of programs, particularly JavaScript The creation of interactive features The inclusion of multimedia objects such as video and sound files If you use a LATEX implementation that produces the typeset file in PDF format, then you already have the file you need. If not, the typeset file will be a DVI file,
13.2
Hyperlinks for LATEX
319
and you have to convert it to PDF format. Usually, this is a two step process. First you convert the DVI file to PostScript and then from PostScript to PDF. On a Mac, just click on a PostScript file and it is converted to PDF. If you have a PostScript printer, you can “print to file” to create the PostScript version. To convert from Postscript to PDF, you need special software such as Adobe Acrobat Professional—free utilities to do this are also available for all operating systems. When downloaded, a PDF file can be read and printed using Adobe Reader (or some other PDF viewer). Adobe Reader is available from Adobe free of charge for a PC, Mac, and some UNIX variants (see Section E.1 on how to get it). PDF files can also be used to make legacy documents available on the Internet. For instance, if you go to my home page, http://www.maths.umanitoba.ca/homepages/gratzer/ and click on Mathematical articles, then 198089, in the bottom frame you will find entry 102, which links to a PDF file. I created that PDF file by scanning the pages of the original article, converting them to PDF files (with Adobe Photoshop), and finally stringing them together into a single document (using Adobe Acrobat Professional). The scanned pages totalled 32 MB, the PDF file is 320 KB. The printed version of the PDF file is somewhat lower in quality than the original, but it is still quite satisfactory. See my article [32] on some practical pointers about scanning and PDF files.
13.1.3 Hyperlinks With Adobe Acrobat Professional you can place hyperlinks in PDF documents. Clicking on a hyperlink, you jump to another location in the same document, to an electronic document, or to a Web site. For instance, in the table of contents, you can put a hyperlink to Chapter 3, so that clicking on it takes you to Chapter 3. Adobe Acrobat’s help system has ample information on how to set up links.
13.2
Hyperlinks for LATEX It is tedious to set hyperlinks one at a time in your PDF file. Would it not be nice if hyperlinks corresponding to crossreferences were set automatically? For instance, clicking on Lemma 6 in
This follows from Lemma 6 and the relevant definitions. would cause the display to jump to the page containing Lemma 6. Sebastian Rahtz’s hyperref package (maintained now by Heiko Oberdiek) does just that (see Section E.1 on how to get it).
320
13.2.1
Chapter 13 PDF documents
Using hyperref You invoke the hyperref package with the command \usepackage{hyperref} as the last \usepackage line in the preamble of your LATEX document. If this does not do the job, try this format, specifying the printer driver: \usepackage[driver ]{hyperref} The driver is one of hypertex, dvips, dvipsone, ps2pdf, tex4ht, pdftex, dvipdf, dvipdfm, dvipdfmx, dviwindo, vtex, textures. If none of these work, you are out of luck. Figure 13.1 shows a page fragment from a mathematical article with hyperlinks to some sections, theorems, and citations automatically created by hypertex.
Figure 13.1: The hyperref package with the \autoref command. See Section 13.2.4 for the autoref command. To see how hypertex works, look up the sampartref.tex article in the samples folder. It is the sample article enhanced with the hyperref package. The article uses the hyperref options pagebackref,colorlinks,bookmarks=true See Section 13.2.2 for the pagebackref and colorlinks options. The third option, bookmarks=true, is discussed in Section 13.2.3. Copy sampartref.tex into the work folder and typeset it twice. The PDF file created for you has some of the hyperref features, but not all. Open the pdf file with Adobe Reader. Look at the left pane. Bookmarks is a table of contents of the article, with links to the named sections. Pages is a thumbnail sketches of the pages, with links to them.
13.2.2 backref and colorlinks A useful addition to hyperref is David Carlisle’s backref package. It is invoked as an option of hyperref: \usepackage[backref]{hyperref}
13.2
Hyperlinks for LATEX
321
The items in your bibliography will be followed by a list of sections in which the bibliographic reference is cited. Each number printed after the cited reference becomes a hyperlink to the relevant section. Alternatively, you can use the pagebackref option, which produces a list of page numbers. Figure 13.2 shows a page fragment from a bibliography displaying lists of section numbers. backref can be used to check if all items in the bibliography have actually been referenced in the article. Any reference that has not been cited does not have a page listed.
Figure 13.2: The hyperref package with the backref option. Another popular option is colorlinks, which colors the text of the links instead of underlining them.
13.2.3 Bookmarks An important navigational feature of Acrobat is the ability to set and use bookmarks. If you choose View>Navigation Tabs>Bookmarks in Adobe Reader, the navigation pane opens up showing the bookmarks. For a sophisticated example, in Adobe Reader, view Adobe Reader Help, the menu item Help>Adobe Reader Help.... The bookmarks form a table of contents. Clicking on a chapter title bookmark displays the first page of the chapter or its table of contents in the main pane. Click on the triangle next to the left of the chapter title bookmark. It opens up the chapter title to show the sections within the chapter, which are links to the sections. The hyperref package option bookmarks=true makes bookmarks from the sectioning commands of the LATEX document, thereby producing a table of contents even if the document had none. You can invoke all these options together: \usepackage[backref,colorlinks,bookmarks=true]{hyperref} Typesetting your LATEX document with the bookmarks=true option produces an out file, which contains entries such as \BOOKMARK [1][]{section.1}{1. Introduction}{} \BOOKMARK [1][]{section*.2}{References}{} Once you have produced the final version of your document, you should edit this file to make sure that it contains no LATEX code. Math formulas in titles create havoc. So do
322
Chapter 13 PDF documents accented characters. hyperref does its best to convert internal encodings for accented characters to the encoding used by Acrobat Reader, but it is still best to avoid them. Once this file has been edited, add the line \let\WriteBookmarks\relax at the start of the file to prevent it from being overwritten.
13.2.4
Additional commands The hyperref package has dozens of commands and parameters, but we will discuss only four more commands. Preventing links If you do not want a \ref or \pageref command to appear as a link, you can use their *ed forms, \ref* and \pageref*. Long links An often heard complaint is that in the link Theorem 6, only the 6 can be clicked to activate the link, and it is too short. hyperref provides the \autoref command to help out. Instead of Theorem~\ref{T:new} you can simply type \autoref{T:new} and hyperref will provide the word Theorem so that the link becomes Theorem 6. The names supported by the \autoref command are listed in Table 13.2.4. For my own use, I redefine: \renewcommand{\chaptername}{Chapter} \renewcommand{\sectionname}{Section} \renewcommand{\subsectionname}{Section} \renewcommand{\subsubsectionname}{Section} External links External links can be links to websites or other files that are located on the Internet. Use the \href{address }{text }
13.2
323
Hyperlinks for LATEX Command
Meaning
\figurename \tablename \partname \appendixname \equationname \Itemname \chaptername \sectionname \subsectionname \subsubsectionname \paragraphname \Hfootnotename \AMSname \theoremname
Figure Table Part Appendix Equation item chapter section subsection subsubsection paragraph footnote Equation Theorem
Table 13.1: Redefinable names supported by \autoref. command to typeset text and make it into a link to the Web address (URL). For instance, in your references, you may have Robert Miner and Jeff Schaefer, \emph{Gentle intoduction to MathML.}\\ \href{http://www.webeq.com/mathml/gitmml/} {http://www.webeq.com/mathml/gitmml/} Then the last line of the address becomes a link and clicking on it takes you to the Web site. As an even fancier example, note the top matter command \urladdr (see Section 11.2.2) in sampartref.tex: \urladdr{\href{http://math.uwinnebago.edu/menuhin/} http://math.uwinnebago.edu/homepages/menuhin/} Then, as part of Menuhin’s address, you will find http://math.uwinnebago.edu/menuhin/ Now clicking on the Web address will link to his Web page. hyperref, of course, offers a lot more than I have presented here. For more detail, see the user manual and The LATEX Web Companion [18].
CHAPTER
14 Presentations
In Section 4.4, we describe how a presentation is a PDF file that you open with Adobe Reader. You can put it in full screen mode1 (View>Full screen), and then project the presentation one page at a time by pressing the space bar or the arrow keys. Remember overhead transparencies? If we want to see half of what is on the transparency, we cover up the bottom part so that only the top part is projected. This way we have control over what the audience sees and when. We sometimes used overlays: placing another transparency on top of the projected one to modify it by adding text or graphics. In this chapter, we discuss Till Tantau’s beamer package to help you prepare presentations with overlays and with stunning visual effects. beamer relies on other packages such as the hyperref package (see Section 13.2) to establish links, Till Tantau’s Portable Graphics Format package for creating graphics, Uwe Kern’s xcolor package for coloring, the AMS packages for formatting math formulas and defining declarations, and some others. The documentation for these packages runs to about a thousand pages. The good news is that you can use beamer “out of the box”. You only have to learn about 20 1 If you have Adobe Acrobat Pro, open File>Document Properties and check mark Full Screen Mode. Then the PDF document automatically opens in full screen mode.
326
Chapter 14
Presentations
commands—this is more than the one new command we had to learn in Section 4.4 but still an easily manageable task. So we set ourselves in this chapter a modest goal, using beamer “out of the box”. It is amazing how much you can achieve with a small investment of your time. Getting started First, you need to install beamer. If you installed MikTeX for your PC or TeXShop for your Mac as described in Appendix A, beamer is already installed for you. Otherwise, consult Section E.1 on how to get it. Second, the tex file created for beamer has to be turned into a PDF file to make it into a presentation. For MikTeX, click on the PDF TeXify icon and for TeXShop, just click on Typeset. For other TEX installations, check the user manual on how to create a PDF output.
14.1
Quick and dirty beamer We convert the article intrart.tex (in the samples folder) to a beamer presentation. We will remove some commands that are appropriate for an article but not for a presentation and add some commands—such as the frame commands—that are specific to presentations. This will not produce a very good presentation. Nevertheless, the conversion is a really quick introduction to some basic beamer concepts.
14.1.1
First changes Open intrart.tex, save it as quickbeamer.tex in the work folder. The converted tex version and the presentation quickbeamer.pdf are both in the samples folder. Make the following changes in the preamble and top matter: 1. Change the first line to % Introductory beamer presentation: quickbeamer.tex 2. Change the documentclass to beamer. 3. Delete the six \usepackage and \newtheorem lines—beamer loads the necessary packages and defines these declarations. 4. Change the \address to \institute—this is the beamer command for address. 5. Delete the abstract environment—this is not needed for the presentation. Here is the new version of the preamble and top matter. %Introductory beamer presentation: quickbeamer.tex \documentclass{beamer} \begin{document} \title{A construction of completesimple\\
14.1
Quick and dirty beamer
327
distributive lattices} \author{George~A. Menuhin} \institute{Computer Science Department\\ University of Winnebago\\ Winnebago, MN 53714} \date{March 15, 2006} \maketitle
14.1.2 Changes in the body 1. Delete the notation and proof environments, but not the contents, that is, delete the four lines \begin{notation} \end{notation} \begin{proof} \end{proof} Both environments could theoretically stay, but the notation environment is not needed since in the next step we put Notation in the frame title. The proof environment is not suitable for presentations because it can only be used within a frame (see Section 5.3.2). 2. Cut the presentation into frames (pages, transparencies, foils) with the frame environments. After each \begin{frame} we put a \frametitle command. The argument of the command is the “title” for the frame, displayed prominently at the top of the display. It would be tedious to give you precise instructions on how to do this, instead refer to the quickbeamer.tex document (in the samples folder) for all the frame environments and \frametitle commands we have added. 3. Cut out the figure environment, except for the line \centering\includegraphics{products} which should be moved to follow the \frametitle{Illustrating the construction} line and accordingly delete (see Figure~\ref{Fi:products}). Now copy over the illustration products from the samples folder to the work folder and typeset. That’s it, enjoy your first presentation.
328
14.1.3
Chapter 14
Presentations
Making things prettier Now you make some small changes to quickbeamer.tex to utilize beamer’s power for wonderful effects. Changes 1 and 3 are quite dramatic. Save quickbeamer.tex with the name quickbeamer1.tex in the work folder. The edited version,quickbeamer1.tex, is in the samples folder along with the presentation quickbeamer1.pdf. 1. Add \usetheme{Berkeley} after the documentclass line. 2. Change \maketitle to \begin{frame} \titlepage \end{frame} Make sure that the last (sub)section is followed by a frame, otherwise it will be missing from the table of contents. 3. Add this frame after the titlepage frame: \begin{frame} \frametitle{Outline} \tableofcontents[pausesections] \end{frame} This creates a table of contents frame, with the section titles appearing one at a time. 4. Replace all instances of {equation} by {equation*}. In a presentation a reference to another frame is not recommended so equations should not be numbered. You might as well delete all the \label commands since these are not needed either. 5. In the second to last frame there are two references to equation numbers. Replace the text the congruence \eqref{E:cong2} with the congruence, and also replace \eqref{E:comp} with the penultimate equation, or similar. 6. Change the bibliographic reference to See also Ernest T. Moynahan, 1957. Turning quickbeamer1.tex into a PDF file will get you a much prettier presentation. The first four pages of the new presentation are displayed in Figures 14.1 and 14.2—unfortunately, without the pretty colors.
14.1.4
Adjusting the navigation Looking at Figures 14.1 and 14.2, we see that the Berkeley theme turns the sidebar into a navigation device and the section titles produce the table of contents (the Outline frame). Remember to typeset a few times! But a number of problems come to light.
14.1
329
Quick and dirty beamer
A construction of completesimple distributive lattices
A construction of completesimple distributive lattices
George A. Menuhin Introduction The Π∗ construction
George A. Menuhin Computer Science Department University of Winnebago Winnebago, MN 53714
March 15, 2006
Outline A construction of completesimple distributive lattices George A. Menuhin Introduction The Π∗ construction
1 Introduction
Figure 14.1: quickbeamer1 presentation, pages 1 and 2.
330
Chapter 14
Presentations
Outline A construction of completesimple distributive lattices George A. Menuhin Introduction The Π∗ construction
1 Introduction
2 The Π∗ construction
Introduction A construction of completesimple distributive lattices George A. Menuhin Introduction
In this note, we prove the following result:
The Π∗ construction
Theorem There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations.
Figure 14.2: quickbeamer1 presentation, pages 3 and 4.
14.1
Quick and dirty beamer
331
1. The title of the presentation is too long for the sidebar, so is the title of the second section. 2. There is no need to repeat the author’s name in the sidebar. 3. It was natural for the article intrart.tex to have only two sections. But sections play a different role in a beamer presentation, they are navigation tools. The sidebar lists all the sections. It also highlights the section we are in. Moreover, by clicking on the name of a section, the presentation jumps there. To correct these deficiencies, save the file quickbeamer1.tex as quickbeamer2.tex in the work folder. The edited version is in the samples folder along with the PDF file. 1. Change the \title command to \title[Completesimple distributive lattices]% {A construction of completesimple\\ distributive lattices} and the second \section command to \section[Construction]{The $\Pi^{*}$ construction} The bracketed parts are the short versions used in the sidebar. 2. Change the \author command to \author[]{George~A. Menuhin} The short version of the author command is blank, so the author’s name will not be displayed in the sidebar. 3. Add the command \section[Second result]{The second result} before the frame of the same title and \section{Proof} before the proof. We even add \section{References} before the frame of the same name. Figure 14.3 shows page 7 of the quickbeamer2 presentation—this corresponds to page 4 of the quickbeamer1 presentation; the Outline accounts for the difference. Note how all the deficiencies listed above have been corrected. Compare page 7 of this presentation with the Berkeley theme in Figure 14.3 and with the Warsaw theme in Figure 14.4. Themes are discussed in Section 14.5.
332
Chapter 14
Presentations
Introduction Completesimple distributive lattices
Introduction Construction Second result Proof References
In this note, we prove the following result: Theorem There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations.
Figure 14.3: quickbeamer2 presentation, page 7 with Berkeley theme. Introduction Construction Second result Proof References
Introduction
In this note, we prove the following result: Theorem There exists an infinite complete distributive lattice K with only the two trivial complete congruence relations.
Completesimple distributive lattices
Figure 14.4: quickbeamer2 presentation, page 7 with Warsaw theme.
14.2
Baby beamers
14.2
333
Baby beamers In the previous discussions you may have noticed two interesting features. First, the Outline frame (table of contents) created two pages in the quickbeamer1 presentation and five pages in the quickbeamer2 presentation. We discuss this in some detail now with the babybeamer presentations. You can find all the babybeamer presentations as tex and PDF files in the samples folder. Second, the sidebar shows some links. More about this in Section 14.2.7.
14.2.1 Overlays The outline frame of the quickbeamer2 presentation created five pages in the PDF file. Observe how each page, from the second on, completely overlaps the previous one, making it appear that the previous one stayed put and an additional line is displayed “on top of it”. In beamer terminology these pages are overlays or slides. The five overlays will be referenced as overlay 1, . . . , overlay 5. A single frame may create one or many overlays. The subsequent sections discuss many more variants. beamer has many commands creating overlays. We start with some examples of \pause, then \only, and \onslide. We introduce overlays with some presentations. The first, babybeamer1, introduces the \pause command to create overlays. % babybeamer1 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Some background} We start our discussion with some concepts. \pause The first concept we introduce originates with Erd\H os. \end{frame} \end{document} produces the presentation of Figure 14.5.
Rule
The \pause command 1. A frame may have many \pause commands. 2. The \pause command cannot be given in an AMS multiline math environment.
334
Chapter 14
Presentations
Some background
We start our discussion with some concepts.
Some background
We start our discussion with some concepts. The first concept we introduce originates with Erd˝os.
Figure 14.5: babybeamer1 presentation. You move past a \pause command the same way as you get to the next frame, by pressing the space bar or the forward arrow key. Using the \pause commands you can create many overlays, each containing a little more material on the overlays. If this is all you need, skip to Section 14.2.7, you do not need the more detailed discussion of overlays in the next few pages. We could have coded the same presentation with the \only command: % babybeamer2 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Some background} \only{We start our discussion with some concepts.} \only{The first concept we introduce originates with Erd\H os.}
14.2
Baby beamers
335
\end{frame} \end{document} This presentation is slightly different from babybeamer1. Overlay 1 ignores the second \only command and displays the line as appropriate to display one line. Overlay 2 displays the two lines as appropriate to display two lines. As a result, the first line moves slightly up when passing from overlay 1 to overlay 2. The argument of the \only command is typeset only on the overlays specified. On the other overlays, it is ignored. If instead of the \only command you use the \onslide command (on slide, get it?), as in \onslide{We start our discussion with some concepts.} \onslide{The first concept we introduce originates with Erd\H os.} then the first line of overlay 2 completely overlaps the first line of overlay 1, so the first line seems to stay put. The argument of the \onslide command is typeset on the overlays specified and on the other overlays it is typeset but invisible. This is the behavior you would want most often, but you may find that sometimes you prefer \only.
14.2.2 Understanding overlays In Section 14.2.1 we introduced overlays, probably the most important new concept for presentations. LATEX typesets the content of a frame and the typeset material appears on all overlays for the parts of the source (maybe all) not modified by any command with an overlay specification; appears only on the overlays specified and is ignored on the other overlays for the arguments of the \only commands; appears on the overlays specified and is typeset but made invisible on the other overlays for the arguments of the \onslide commands. More on overlay specifications at the end of this section. Here are some illustrations. Example 1 This is a very \only{very, very} important concept. \only{To start the definition \dots} will typeset overlay 1 as
336
Chapter 14
Presentations
This is a very very, very important concept. To start the definition . . . and will typeset overlay 2 as This is a very important concept. To start the definition . . .
Example 2 What is $2+2$? It is \onslide{$4$}. \only{Can you figure it out?} \onslide{I hope you all got it right.} will typeset overlay 1 as What is 2 + 2? It is . Can you figure it out? and will typeset overlay 2 as What is 2 + 2? It is 4. I hope you all got it right. Note that there is room in overlay 1 for the number 4. Example 3 What is $2+2$? \onslide{It is {$4$}.} Can you figure it out? will typeset overlay 1 as What is 2 + 2? Can you figure it out?
14.2
Baby beamers
337
and will typeset overlay 2 as What is 2 + 2? It is 4. Can you figure it out? Note that there is room in overlay 1 for the “missing” second line. Overlay specifications The angle brackets contain the overlay specification. Here are some more examples: means all overlays from 1 to 2, and all overlays from 4 onwards. means all overlays up to 3. means overlays 2, 4, and 6. In the presentation babybeamer2 we have two overlay specifications: and . Maybe, and would be better, so that if you add a third overlay you do not have to change these. The command \pause can only take the simplest overlay specification, a number. \pause takes effect from overlay 3 on. Note that overlay specifications are attached to commands but the overlays created are overlays of the frame in which the commands appear.
14.2.3 More on the \only and \onslide commands The \only and \onslide commands can accomplish everything the \pause command can and a lot more. The basic syntax The syntax of \only is \only{source } where overlay spec is the overlay specification and source is the code typeset by LATEX. A (partial) syntax of \onslide is \onslide{source } With the same syntax you can give overlay specifications to many commands, including \textbf, \textit, \alert—beamer’s alternative to the \emph command— and then the command is in effect only on the overlays specified.
338
Chapter 14
Presentations
% babybeamer3 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Some background} \textbf{We start our discussion with some concepts.} \textbf{The first concept we introduce originates with Erd\H os.} \end{frame} \end{document} So the babybeamer3 presentation (see Figure 14.6) has two overlays, each with two lines of text. On overlay 1 the first line is bold, on overlay 2 the second line is bold.
Some background
We start our discussion with some concepts. The first concept we introduce originates with Erd˝os.
Some background
We start our discussion with some concepts. The first concept we introduce originates with Erd˝ os.
Figure 14.6: babybeamer3 presentation.
14.2
Baby beamers
339
A different syntax The command \only has an alternate syntax: \only{source } So \only{Can you figure it out?} and \only{Can you figure it out?} accomplish the same. With this syntax, you can define your own commands that allow overlay specifications. For instance, using the command \color{blue} defined in Section 14.2.9, you can define the command \newcommand{\myblue}{\only{\color{blue}}} Then \myblue{Some more text} will color the text blue on overlay 2 only.
14.2.4 Lists as overlays Lists may be presented one item at a time, for example the babybeamer4 presentation in Figure 14.7 (in the samples folder) shows the four overlays of a list. R. Padmanabhan appears on the first, R. Padmanabhan and Brian Davey appear on the second, and so on. This is accomplished simply by adding the overlay specification to the item for R. Padmanabhan, the overlay specification to the item for Brian Davey, and so on. % babybeamer4 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Overlaying lists} We introduce our guests: \begin{itemize} \item R. Padmanabhan \item Brian Davey
340
Chapter 14 Overlaying lists
We introduce our guests: ◮
R. Padmanabhan
Overlaying lists
We introduce our guests: ◮
R. Padmanabhan
◮
Brian Davey
Overlaying lists
We introduce our guests: ◮
R. Padmanabhan
◮
Brian Davey
◮
Harry Lakser
Overlaying lists
We introduce our guests: ◮
R. Padmanabhan
◮
Brian Davey
◮
Harry Lakser
◮
Dick Koch
Figure 14.7: babybeamer4 presentation.
Presentations
14.2
Baby beamers
341
\item Harry Lakser \item Dick Koch \end{itemize} \end{frame} \end{document} Such an overlay structure is used so often that beamer has a shorthand for it, []. Here it is in babybeamer5. % babybeamer5 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Overlaying lists} We introduce our guests: \begin{itemize}[] \item R. Padmanabhan \item Brian Davey \item Harry Lakser \item Dick Koch \end{itemize} \end{frame} \end{document} This shorthand allows adding and reordering items without having to change overlay specifications. Of course, if you do not want the items to appear in sequence, you have to use overlay specifications.
14.2.5 Out of sequence overlays We now present an example of “out of sequence overlays”. Look at Figure 14.8. I want to make this part of my presentation. First, I want to show the theorem, then illustrate it with the diagram at the bottom. Finally, I present the proof in the middle. So I need three overlays. The theorem is on all three overlays, 1, 2, 3. Its illustration is on overlays 2 and 3, leaving room for the proof that appears only on overlay 3. This is an example of “out of sequence overlays”. We code this in babybeamer6 (in the samples folder). Since declarations, proofs, and the \includegraphics command may all have overlay specifications, this seems easy to accomplish.
342
Chapter 14
Presentations
% babybeamer6 presentation, first try \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Overlaying declarations and graphics} \begin{theorem} Every finite distributive lattice can be embedded in a boolean lattice. \end{theorem} \begin{proof} Use joinirreducible elements. \end{proof} \includegraphics{cube} \end{frame} \end{document} This does not work too well. On overlay 1 the theorem appears in the middle and then it jumps up to make room for the illustration. This is the same problem we encountered in the babybeamer2 presentation in Section 14.2.1 and the solution is also the same, the use of the \onslide command. Replace the line \includegraphics{cube} with \onslide{\includegraphics{cube}}
Theorem
Every finite distributive lattice can be embedded in a boolean lattice.
Proof.
Use joinirreducible elements.
Figure 14.8: The slide to represent.
14.2
Baby beamers
343
14.2.6 Blocks and overlays You can think of a theorem in beamer as the contents of the theorem environment with a heading and, optionally, with an overlay specification, and with most themes— see Section 14.5—colorful visual highlighting, see Figures 14.3 and 14.4. beamer provides the block environment that works the same way except that you name the block. The (partial) syntax of the block environment is \begin{block}{title } source \end{block} Blocks are shaped as theorems. If there is no title, you still need the braces. The overlay specification is optional. As an example, save babybeamer6.tex as babybeamer6block.tex in the work folder (also in the samples folder along with the PDF file) and replace the theorem environment with \begin{block}{Theorem} Every finite distributive lattice can be embedded in a boolean lattice. \end{block} If you want a block of LATEX code with an overlay specification but with no title and no visual highlighting, use one of the commands, \onslide and \only.
14.2.7 Links A presentation is a PDF file, so it is not surprising that you can set links of various types in a beamer presentation. Just as the hyperref package helps us with hyperlinks in a PDF file (see Section 13.2), the beamer package allows us to conveniently set links in a presentation. Some links are automatically provided. If you look closer at Figures 14.1 and 14.2, you see that the section titles are shown in the sidebar. In fact, the sidebar is a navigation bar. First, it shows which section you are in. Second, clicking on a section title takes you to that section. Creating a link is a twostep process. 1. Name the place you want to link to. 2. Create a button with the property that clicking on it jumps you to the designated place. To illustrate this process, we modify the presentation babybeamer4. Open the file babybeamer4.tex and save it as babybeamer7.tex in the work folder (the edited version is in the samples folder along with the PDF file).
344
Chapter 14
Presentations
1. Name the frame you want to link to by adding a label to the \begin{frame} line. In babybeamer7, add a label to the frame fourguests: \begin{frame}[label=fourguests] Labels of frames are also useful for selective typesetting of your presentation, see Section 14.6. 2. Add the following line to babybeamer7: \hyperlink{fourguests}% {\beamergotobutton{Jump to third guest}} This creates a link to the third overlay of the frame named fourguests, and creates a button, with the text Jump to third guest. Clicking on this button will jump to the third overlay of the frame fourguests. 3. To add variety to linking, include a new first frame: \begin{frame} \frametitle{First frame with a button} Button example Jumping to an overlay of a different frame \bigskip \hyperlink{fourguests}% {\beamergotobutton{Jump to third guest}} \end{frame} which has a button for jumping to the third overlay of the fourguests frame. 4. We also add a new third frame. \begin{frame} \frametitle{Third frame with a button} Button example Jumping to another frame \bigskip \hyperlink{fourguests}% {\beamergotobutton{Jump to guest list}} \end{frame} with a button, with the text Jump to guest list. Clicking on this button will jump to the second frame, overlay not specified (defaults to 1).
14.2
Baby beamers
345
5. Add a fourth frame, \begin{frame} \frametitle{Hidden link} \hyperlink{fourguests}{Jumping to the guest list} \end{frame} introducing another version of the \hyperlink command: \hyperlink{fourguests}{Jumping to the guest list} which typesets the second argument as regular text, making it an invisible link. However, you may notice that the cursor changes when it hovers over the link. For instance, you may want to link the use of a concept to its earlier definition, where you also need a button for the return jump. Here is babybeamer7: % babybeamer7 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{First frame with a button} Button example Jumping to an overlay of a different frame \bigskip \hyperlink{fourguests}% {\beamergotobutton{Jump to third guest}} \end{frame} \begin{frame}[label=fourguests] \frametitle{Overlaying lists} We introduce our guests: \begin{itemize} \item R. Padmanabhan \item Brian Davey \item Harry Lakser \item Dick Koch \end{itemize}
346
Chapter 14 First frame with a button
Button example Jumping to an overlay of a different frame Jump to third guest
Overlaying lists
We introduce our guests: ◮
R. Padmanabhan
Jump to third guest
Overlaying lists
We introduce our guests: ◮
R. Padmanabhan
◮
Brian Davey
Jump to third guest
Third frame with a button
Button example Jumping to another frame Jump to guest list
Figure 14.9: babybeamer7 presentation.
Presentations
14.2
Baby beamers
347
\hyperlink{fourguests}% {\beamergotobutton{Jump to third guest}} \end{frame} \begin{frame} \frametitle{Third frame with a button} Button example Jumping to another frame \bigskip \hyperlink{fourguests}% {\beamergotobutton{Jump to guest list}} \end{frame} \begin{frame} \frametitle{Hidden link} \hyperlink{fourguests}{Jumping to the guest list} \end{frame} \end{document} Figure 14.9 shows all these buttons. We do not show overlays 3 and 4 of frame 2 and frame 4, where the button is invisible.
14.2.8 Columns Often, it is useful to put the display into columns. A simple illustration is given in babybeamer8: % babybeamer8 presentation \documentclass{beamer} \begin{document} \begin{frame} \frametitle{Columns, top alignment} \begin{columns}[t] \begin{column}{2in} Is it true that there is no new result on the Congruence Lattice Characterization Problem? \end{column} \begin{column}{2in}
348
Chapter 14
Presentations
F. Wehrung found a distributive algebraic lattice that cannot be represented as the congruence lattice of a lattice. \end{column} \end{columns} \end{frame} \end{document} The environment is columns. It has an optional argument for alignment, t for top, c for center, and b for bottom. The columns, usually two, are both in the column environment; the width of the column is in the argument; it can be given as a measurement—2in in the example—or relative to the width of the whole frame as 0.4\textwidth. Figure 14.10 shows the babybeamer8 presentation. Columns, top alignment
Is it true that there is no new result on the Congruence Lattice Characterization Problem?
F. Wehrung found a distributive algebraic lattice that cannot be represented as the congruence lattice of a lattice.
Figure 14.10: babybeamer8 presentation.
14.2.9
Coloring LATEX’s job is to produce articles and books that contain text, math formulas, and graphics. Such publications—with the exception of textbooks—cannot afford color printing. Presentations are different. If you prepare a color presentation, it will project in color. Nevertheless, the color commands are of limited use even for presentations. You probably use the color scheme of the chosen theme (see Section 14.5), and have limited opportunity to color things yourself. If you do, be very careful, too much color distracts from the presentation but judicious use of color—say, for highlighting a word or phrase—may be very effective. beamer uses the sophisticated xcolor package of Uwe Kern. It colors by specifying the color model: rgb (red, green, blue), or cmyk (cyan, magenta, yellow, black), or gray (black and white)—there are many more models to choose from—and how much
14.2
Baby beamers
349
of each color you want to mix. So \color[rgb]{0,1,0} paints everything—within its scope—green. You can color some text green with the command \textcolor[rgb]{0,1,0}{This text is green.} There are seventeen predefined colors: red, green, blue, cyan, magenta, yellow, orange, violet, purple, brown, pink, olive, black, darkgray, gray, lightgray, and white. With the proper options, there are hundreds more. So the previous command could also be given as \textcolor{green}{This text is green.} or as {\color{green}This text is green.} To pretty thing up, you can use \colorbox{green}{Green box}, which puts the argument in a green box and \fcolorbox{red}{green}{Green box}, which also adds a red frame. xcolor is automatically loaded by beamer. To make sure that xcolor is loaded with the options desired, you have to include these options in the preamble in the \documentclass{beamer} line. For instance, to have the dvipsnam option for xcolor, invoke beamer with \documentclass[xcolor=dvipsnam]{beamer} You can also mix predefined colors: {\color{green!40!yellow} This text is of what color?} which sets the text 40% green and 60% yellow. There are commands for defining colors and color sets, as well as for coloring the background, frames, and hyperlinks (see Uwe Kern, Extending LATEX’s color facilities: the xcolor package [36]). Here is a simple illustration: % babybeamer9 presentation \documentclass{beamer} \begin{document} \setbeamercolor{normal text}{bg=yellow!15} \begin{frame} \frametitle{Colors} \begin{columns}[t] \begin{column}{2in}
350
Chapter 14
Presentations
{\color{red}Is it true that there is no new result on the Congruence Lattice Characterization Problem?} \end{column} \begin{column}{2in} {\color{green}F. Wehrung found a distributive algebraic lattice that cannot be represented as the congruence lattice of a lattice.} \end{column} \end{columns} \end{frame} \setbeamercolor{normal text}{bg=green!15} \begin{frame} \frametitle{Colors fading out} We introduce our guests: \begin{itemize} \item {\color{red}R. Padmanabhan} \item {\color{red!60!white}Brian Davey} \item {\color{red!40!white}Harry Lakser} \item {\color{red!20!white}Dick Koch} \end{itemize} \end{frame} \end{document} The command \setbeamercolor{normal text}{bg=yellow!15} sets the background color to light (15%) yellow. In the first column, the text is red, in the second, green. Set the foreground with fg=. The background of the second frame is light green. The four participants are in lighter and lighter shades of red. Figure 14.11 shows the frames of the babybeamer9 presentation—unfortunately, not in color.
14.3
The structure of a presentation The structure of your presentation is, by and large, determined by the sectioning commands: \section and \subsection. For a very long lecture there may also be \part commands. The argument of any of these commands may have a short version for the navigational side bar (see Section 14.1.4). The sectioning commands used in a beamer presentation look the same as they do for articles and books, but they play a different role. They do not display a section
14.3
351
The structure of a presentation Colors
Is it true that there is no new result on the Congruence Lattice Characterization Problem?
F. Wehrung found a distributive algebraic lattice that cannot be represented as the congruence lattice of a lattice.
Colors fading out
We introduce our guests: ◮
R. Padmanabhan
◮
Brian Davey
◮
Harry Lakser
◮
Dick Koch
Figure 14.11: babybeamer9 presentation. title, but they add an entry to the table of contents. They also act as place markers in the sense that if you click on the title of a section in a navigation bar, then you will jump to the frame following the section command.
Rule
Sectioning commands 1. Sectioning commands can only be placed between frames. 2. There must be a frame following the last sectioning command. 3. For a long (sub)section title, use \breakhere to break a line. 4. The optional short versions are for the navigation bar. These are illustrated with beamerstructure1, see Figure 14.12. The line \tableofcontents[pausesections, pausesubsections]
352
Chapter 14
Presentations
Outline
Sec1 Sec1 Subsec1 Sec1 Subsec2 Sec1 Subsec3
Sec2
1 Section 1
Section 1 – Subsection 1 Section 1 – Subsection 2
Outline
Sec1 Sec1 Subsec1 Sec1 Subsec2 Sec1 Subsec3
Sec2
1 Section 1
Section 1 – Subsection 1 Section 1 – Subsection 2 Section 1 – Subsection 3
2 Section 2
Figure 14.12: beamerstructure1 presentation, pages 3 and 5. causes the table of contents to appear a line at a time. This command may also be used without an option or only with one, pausesections. The second page shown in Figure 14.12 is the table of contents. The page is about half filled with only five listed items, so no more than 10 sections and subsections would fit. There should be fewer. % beamerstructure1 presentation \documentclass{beamer} \usetheme{Berkeley} \begin{document} \begin{frame} \frametitle{Outline} \tableofcontents[pausesections, pausesubsections] \end{frame}
14.3
The structure of a presentation \section[Sec1]{Section 1} \begin{frame} \frametitle{Section 1} Text of Section 1 \end{frame} \subsection[Sec1 Subsec1]{Section 1  Subsection 1} \begin{frame} \frametitle{Section 1\\Subsection 1} Text of Section 1, Subsection 1 \end{frame} \subsection[Sec1 Subsec2]{Section 1  Subsection 2} \begin{frame} \frametitle{Section 1\\Subsection 2} Text of Section 1, Subsection 2 \end{frame} \subsection[Sec1 Subsec3]{Section 1  Subsection 3} \begin{frame} \frametitle{Section 1\\Subsection 3} Text of Section 1, Subsection 3 \end{frame} \section[Sec2]{Section 2} \begin{frame} \frametitle{Section 2} Text of Section 2 \end{frame} \end{document}
353
354
14.3.1
Chapter 14
Presentations
Longer presentations Longer presentations may need parts and a more complicated table of contents. I will not discuss these topics, but the presentation beamerstructure2 (in the samples folder) illustrates the use of parts and some other features. I added some comments to point these out. See Figure 14.13 for two sample pages of this presentation. Outline
Sectionally complemented chopped lattices George Gr¨atzer1
Harry Lakser1 1 University 2 Brandon
Michael Roddy2
of Manitoba University
Conference on Lattice Theory, 2006
George Gr¨ atzer, Harry Lakser, Michael Roddy
Sectionally complemented chopped lattices
Chopped lattices Ideals and congruences
Part I Outline
1
Chopped lattices
2
Ideals and congruences
George Gr¨ atzer, Harry Lakser, Michael Roddy
Sectionally complemented chopped lattices
Figure 14.13: beamerstructure2 presentation, pages 1 and 10.
14.3.2
Navigation symbols The more structure you have in a presentation, the more you may appreciate the navigation icons shown by default on each page in the last line on the right. The icons are: the slide the frame the section
14.4
355
Notes the presentation icons each surrounded by a left and a right arrow the appendix the back and forward icons (circular arrows) the search icon (a magnifying glass)
If you decide not to have them, as in the presentation beamerstructure2, then give the following command in the preamble: \setbeamertemplate{navigation symbols}{}
14.4
Notes You can place notes in your presentation to remind yourself of what you want to say in addition to what is being projected. A note is placed in the presentation as the argument of the \note command, as in \note{This is really difficult to compute.} By default, notes are not shown in the presentation. If you invoke beamer with \documentclass[notes=show]{beamer} then the notes pages are included. The command \documentclass[notes=show, trans]{beamer} produces transparencies with notes, and \documentclass[notes=only]{beamer} produces only the note pages, one note page for every overlay of a frame with a note. To avoid this, print the output of \documentclass[trans, notes=only]{beamer} In addition to these examples, all the notes placed in a single frame are collected together on one note page. And a note between frames becomes a page on its own. beamer does an excellent job of producing notes pages, for example, see Figure 14.14. In the upperleft corner, it displays precisely where we are in the structure of the presentation. The upperright corner shows a small picture of the page to which the notes are attached.
356
Chapter 14
20060423
Sectionally complemented chopped lattices
Presentations
The 1960 Theorem
Background The 1960 sectional complement The 1960 Theorem
Theorem (Gr¨ atzerSchmidt 1960) Let M be the chopped lattice constructed from a finite order P by merging. Then the ideal lattice of M is a sectionally complemented lattice.
And the congruence lattice of L is isomorphic to a finite distributive lattice D whose joinirreducible elements form an order isomorphic to P. What we are discussing here is commonly called the Dilworth Theorem, see the book Congruences of a finite lattice, Birkh¨auser Boston, 2005. Figure 14.14: A note page.
14.5
Themes If you look carefully at Figures 14.1, 14.2, and 14.13 (even better, if you look at the PDF files of these presentations), you see how every little detail of the presentation is provided by beamer. Figures 14.1 and 14.2 show a presentation style dominated by a dark blue headline and left sidebar, a complete navigation bar in the left sidebar, the name of the author(s) and the title repeated in every overlay, section numbers in colored squares, and so on. The presentation in Figure 14.13 has no sidebars, section numbers are in colored circles, the navigation elements are in the headline. The display of lists and theorems (and other similar elements of a presentation) also vary a great deal. You can achieve such detailed control over your presentation by defining all these elements yourself. beamer places dozens of commands at your disposal to make this possible. Or you can use a presentation theme that will do the job for you. The command to name a presentation theme is \usetheme{}. The presentation beamerstructure2 uses the theme Warsaw (see Figure 14.13), so following the document class line type the command \usetheme{Warsaw} quickbeamer1 uses the theme Berkeley (see Figures 14.1 and 14.2) and so does beamerstructure1 (see Figure 14.12). The presentation themes are in the theme subfolder of the themes folder of beamer. As of this writing, there are 26 of them, named after cities: Presentation Themes Without Navigation Bars default, Bergen, Boadilla, Madrid, AnnArbor, CambridgeUS, Pittsburgh, Rochester With a Navigation Bar Antibes, JuanLesPins, Montpellier
14.5
Themes
357
With a table of contents Sidebar Berkeley, PaloAlto, Goettingen, Marburg, Hannover With Mini Frame Navigation Berlin, Ilmenau, Dresden, Darmstadt, Frankfurt, Singapore, Szeged With Section and Subsection Table Copenhagen, Luebeck, Malmoe, Warsaw How do you choose a presentation theme? After the presentation is finished, try out the various themes. Ask yourself: Do sidebars take too much room away from my illustrations? Do stronger colors add to the presentation or do they distract? Do I want to use a navigation bar? Answering these questions will narrow your choice. The presentation theme defines all the colors, but you can alter them with the command \usecolortheme{}. You have a choice of albatross, beetle, crane, fly, and seagull. For instance, \usetheme{Warsaw} \usecolortheme{seagull} is a gray version of the Warsaw theme, appropriate for printing in blackandwhite. In addition, you can further modify the “inner elements”, such as blocks, with \usecolortheme{lily} or orchid, or rose. You can modify the “outer elements”, such as headlines and sidebars, with \usecolortheme{whale} or seahorse, or dolphin. So you can have, for instance, \usetheme{Warsaw} \usecolortheme{lily} \usecolortheme{whale} This gives you 45 “out of the box” color schemes. Similarly, font themes can also be specified, modifying the presentation theme, with the command \usefonttheme{}. You have the default and the following options: professionalfonts structurebold structureitalicserif structuresmallcapserif.
358
14.6
Chapter 14
Presentations
Planning your presentation Step 1 As a rule, your presentation is based on one or more of your articles. Collect them in one folder. Resolve naming conventions as necessary. There should be only one Fig1! Step 2 Rewrite the article(s) to sketch out your presentation. The pages correspond to frames. A page should not have too many words, say, no more than 40. Replace your numbered theorems with named theorems. Never reference another page. Have few sections and subsections. Add a table of contents, which is a readable overview of the new article. Step 3 Base the new presentation on a presentation in the samples folder, a sample presentation in beamer’s solution folder, or on one of your own or of a colleague’s older presentations. Turn the pages into frames. Step 4 Design your frames and add frame titles. Completely disregard what we wrote in Section 4.3.2 (the idea behind LATEX is that you should concentrate on what you have to say and let LATEX take care of the visual design). The new principle is: You are completely responsible for the visual appearance of every frame and overlay. This is, of course, in addition to brevity and readability. Do not let LATEX break your lines. Do it with the \\ command and keep words that belong together on the same line. Step 5 Write notes to remind yourself what you want to say in your lecture that is not on the slides. Print the notes for your lecture. Step 6 Build in flexibility. For instance, if you have four examples to illustrate a definition, put each one on a different frame or overlay, and add a link to each that skips the rest of the examples. Depending on your audience’s understanding, show an example or two, and skip the rest. The same way, you may skip proof ideas and even topics. Step 7 Prepare for the worst—the computer system may fail, but projectors seldom do—so print a set of transparencies for your lecture as a backup by invoking the option trans of the documentclass \documentclass[trans]{beamer} To print a handout, use the handout option \documentclass[handout]{beamer} Open the presentation in Acrobat Reader. In Printer/Page Setup... set landscape and 140% magnification. In the Print dialogue box in Layout choose two pages per sheet and print—assuming, of course, that you have a printer offering these options.
14.7
What did I leave out? Since the beamer reference manual is over 200 pages long, it is clear that this chapter covers maybe 10% of it.
14.7
What did I leave out?
359
For most presentations, you won’t even need most of what I have included. If you read Sections 14.1, 14.2.1 and maybe Section 14.3, you should have enough for most math presentations. If you are in other fields, or if you are more ambitious, you may need more. For example, a computer scientist will want program listings in a verbatim environment. This is easy. Start your frame with \begin{frame}{fragile} and then you can use the verbatim environment. If you want to include sounds or movies in your presentation, consult Till Tantau, User’s Guide to the Beamer Class [59]. You can do very simple animation with what we have covered here. This is illustrated with the babybeamer10 presentation (in the samples folder). % babybeamer10 presentation \documentclass{beamer} \begin{document} \begin{frame} \includegraphics{basem31} \includegraphics{basem32} \includegraphics{basem33} \includegraphics{basem34} \end{frame} \end{document} The congruence generated by the dashed red line, see Figure 14.15, spreads in three steps, illustrating an interesting result. The animation is quite effective and instructive. If you want to place such changing pictures lower in a frame, put them in the overprint environment. I would recommend that you read Section 5 of Till Tantau, User’s Guide, which has many good pointers about creating presentations.
360
Chapter 14
Figure 14.15: The four pictures of babybeamer10.
Presentations
CHAPTER
15 Customizing LATEX
In Section D.1.2, we discuss that Donald E. Knuth designed TEX as a platform on which convenient work environments could be built. One such work environment, LATEX, predominates today, and it is indeed convenient. Nevertheless, LATEX is designed for all of us, so it is not surprising that we could improve on it for our personal use. There are many reasons to customize LATEX: Goal 1 to enhance the readability of the source file Goal 2 to make notational and terminological changes easier Goal 3 to redefine names used by LATEX Goal 4 to introduce consistent layouts There are many techniques to accomplish these. Technique 1 Define commands and environments in order to enhance LATEX to meet your particular needs (see Sections 15.1, 15.1.6, and 15.2).
364
Chapter 15
Customizing LATEX
Technique 2 Utilize delimited commands to write LATEX documents in a more readable fashion (see Section 15.1.8). Technique 3 Collect your frequently used commands into a command file (see Section 15.3). Technique 4 Manipulate counters, integers—for instance, equation numbers and section numbers—and length commands, distance measurements—the \voffset command is an example (see Section 15.5). Technique 5 Create customized list environments with the list environment (see Section 15.6). Of course, there are many more reasons to customize and many more techniques to employ. We cover them in detail in this chapter. We dedicate the last section to the pitfalls of customization (see Section 15.7). While the benefits of customization are great, there are many practices to avoid.
15.1
Userdefined commands LATEX provides hundreds of commands. Chances are good, however, that you still have specific needs that are not directly addressed by these commands. By judiciously adding userdefined commands (or macros) you can make your work more productive. Userdefined commands follow the same rules as regular LATEX commands (see Section 5.3.1).
15.1.1
Examples and rules Commands to enhance readability Let us start with a few examples of userdefined commands as shorthand for longer command(s) or text in order to enhance readability of the source file (Goal 1). 1. If you use the \leftarrow command a lot, you could define \newcommand{\larr}{\leftarrow} Then you would only have to type \larr to obtain a left arrow. 2. Instead of \widetilde{a} you could simply type \wtilda after defining \newcommand{\wtilda}{\widetilde{a}} I show you how to define a generalized version of such a command in Section 15.1.2.
15.1
Userdefined commands
365
3. If you want to suppress the ligature in iff (see Section 5.4.6), you would normally have to type if\textcompwordmark f By defining a command \Iff, \newcommand{\Iff}{if\textcompwordmark f} you can type \Iff to get iff. We name this command \Iff because \iff is the symbol ⇐⇒ (see Section B.4). 4. If you use the construct D[2] × D[3] often, you could introduce the \DxD (D times D) command, \newcommand{\DxD}{D^{[2]}\times D^{[3]}} and then type \DxD instead of the longer, and hard to read, version throughout your document—serves also Goal 2. 5. If you want to get a backslash in typewriter style, you would normally have to type (see Section 5.4.4) \texttt{\symbol{92}} Instead, you can introduce the \bsl command, \newcommand{\bsl}{\texttt{\symbol{92}}} and \bsl typesets as \. 6. You can also use commands as a shorthand for text. For instance, if you use the phrase subdirectly irreducible many times in your document, you could define \newcommand{\subdirr}{subdirectly irreducible} \subdirr is now shorthand for subdirectly irreducible, which typesets as subdirectly irreducible.
Tip
With modern editors, the need to have userdefined commands as shorthand is reduced. Most editors have “command completion” or “phrase completion”. For instance, in TeXShop, type the first few letters of a word and hit the escape key. The remaining letters are entered to match the first entry in the completion dictionary. Hitting escape again cycles through all possible completions. To make this feature useful, you have to customize the completion dictionary.
366
Chapter 15
Rule
Customizing LATEX
Userdefined commands 1. Issue the \newcommand command. 2. In braces, type the name of your new command, for example, \subdirr, including the backslash (\). 3. In a second pair of braces, define the command, in this example, subdirectly irreducible. 4. Use the command as \subdirr\ or \subdirr{} before a space, before an alphabetical character as \subdirr{}, and \subdirr otherwise. Examples for Rule 4. For subdirectly irreducible lattice type \subdirr{} lattice or \subdirr\ lattice and not \subdirr lattice. Indeed, typesetting \subdirr lattice results in subdirectly irreduciblelattice. By the first spacing rule, \subdirr lattice is not any better (see Section 5.2.1). If you want subdirectly irreducibles, you must use the \subdirr{} form. Indeed, \subdirr{}s typesets as subdirectly irreducibles. Using new commands It is good practice to place userdefined commands in the preamble of your document or in a command (style) file you load with a \usepackage command (see Section 15.3)— provided that you do not submit to a journal that does not allow this. Then you always know where to look for the command definitions. An exception is a userdefined command that you want to restrict to a part of the document. Delimit the segment with braces and define the userdefined command within those braces (see Section 5.3.2). Instead of a pair of braces, you can use \begingroup and \endgroup, which is easier to see. Section 15.2.5 recommends yet another approach.
Tip If errors occur, isolate the problem. Comment out the userdefined commands and reintroduce them one at a time. LATEX only checks whether the braces match in the command definition. Other mistakes are found only when the command is used. For instance, if you define a command with a spelling error \newcommand{\bfA}{\textf{A}} then at the first use of \bfA you get the error message
15.1
Userdefined commands
367
! Undefined control sequence. \bfA >\textf {A} Note that LATEX is not complaining about \bfA but about the misspelled \textbf command in the definition of \bfA. Be careful not to define a userdefined command with a name that is already in use. If you do, you get an error message such as ! LaTeX Error: Command \larr already defined. To correct the error, replace the command name with a new one. On the other hand, if you need to replace an existing command, you have to redefine it. See Section 15.1.5 for how to do so.
Tip
Use spaces to make your source files more readable, but avoid them in definitions. For example, you may type $D^{ \langle 2 \rangle } + 2 = x^{ \mathbf{a} }$ This may help you see how the braces match, easily identify relations and operations, and so on. Do not add these spaces in command definitions because it may result in unwanted spaces in your typeset document. You may start a new line to increase the readability of a command definition, provided that you terminate the previous line with %. For instance, borrowing an example from page 372: \newcommand{\Xquotphi}[2]{% \dfrac{\varphi \cdot X_{n, #1}}% {\varphi_{#2}\times \varepsilon_{#1}}}
Tip
In the definition of a new command, command declarations need an extra pair of braces (see Section 5.3.3). Say you want to define a command that typesets the warning: Do not redefine this variable! It is very easy to make the following mistake: \newcommand{\Warn}{\em Do not redefine this variable!} \Warn typesets the warning emphasized, but everything that follows the warning is also emphasized (more precisely, until the end of the \Warn command’s scope). Indeed, \Warn is replaced by \em Do not redefine this variable! so the effect of \em goes beyond the sentence to the next closing brace.
368
Chapter 15
Customizing LATEX
The correct definition is \newcommand{\Warn}{{\em Do not redefine this variable!}} Even simpler, you could use a command with an argument \newcommand{\Warn}{\emph{Do not redefine this variable!}} The xspace package Rule 4 (on page 366) is the source of many annoying problems in LATEX. David Carlisle’s xspace package (see Section 12.3.1) helps eliminate such problems. In the preamble, load the package with \usepackage{xspace} Whenever you define a command that may have such problems, add the \xspace command to the definition. For instance, define \subdirr as \newcommand{\subdirr}{subdirectly irreducible\xspace} Then all the following typesets subdirectly irreducible lattice correctly: \subdirr\ lattice \subdirr{} lattice \subdirr lattice Note that \xspace does not add space if followed by a punctuation mark, so to get the lattice is subdirectly irreducible. type the lattice is \subdirr.
Tip
Be careful not to use \xspace twice in a definition. For instance, if you define \newcommand{\tex}{\TeX\xspace} \newcommand{\bibtex}{\textsc{Bib}\kern.1em\tex\xspace}% Bad!!! then \bibtex, followed by a comma
15.1
Userdefined commands
369
typesets as BibTEX , followed by a comma The correct definitions are \newcommand{\tex}{\TeX\xspace} \newcommand{\bibtex}{\textsc{Bib}\kern.1em\TeX\xspace}% Correct! Of course, if you want to get TEXbook, you cannot use \xspace variant definition: \tex. Ensuring math The \ensuremath command is useful for defining commands that work in both text and math mode. Suppose you want to define a command for Dh2i . If you define it as \newcommand{\Dsq}{D^{\langle2\rangle}} then you can use the command in math mode, but not in text mode. If you define it as \newcommand{\Dsq}{$D^{\langle2\rangle}$} then it works in text mode, but not in math mode. Instead, define this command as \newcommand{\Dsq}{\ensuremath{D^{\langle2\rangle}}} Then \Dsq works correctly in both contexts. This example also shows the editorial advantages of userdefined commands. Suppose the referee suggests that you change the notation to D[2] . To carry out the change you only have to change one line: \newcommand{\Dsq}{\ensuremath{D^{[2]}}} It is hard to overemphasize the importance of this example. You may want to change notation because: you found a better notation; your coauthor insists; your article appears in a conference proceedings, and the editor wants to unify the notation; you are reusing the code from this article in another one or in a book, where the notation is different. See also the discussion of the \TextOrMath command on page 311.
370
15.1.2
Chapter 15
Customizing LATEX
Arguments Arguments of userdefined commands work the same way as for LATEX commands (see page 69). Define \newcommand{\fsqAB}{(f^2)^{[[\frac{A^2}{B1}]]}} A2
Then \fsqAB typesets as (f 2 )[[ B−1 ]] in a math formula. If you want to use \fsqAB in math and also by itself in text, define it with \ensuremath, as \newcommand{\fsqAB}{\ensuremath{(f^2)^ {[[\frac{A^2}{B1}]]}}} However, if you use this construct for many functions f , then you may need a generalized command, such as \newcommand{\sqAB}[1]{\ensuremath{(#1^2)^ {[[\frac{A^2}{B1}]]}}} A2
Now \sqAB{g} typesets (g 2 )[[ B−1 ]] . The form of this \newcommand is the same as before, except that after the name of the command in braces, {\sqAB}, we specify the number of arguments in brackets (in this example, [1]). Then we can use #1 in the definition of the command. When the command is invoked, the argument you provide A2
replaces #1 in the definition. Typing $\sqAB{q}$ results in the formula (q 2 )[[ B−1 ]] , A2 [[ B−1 ]]
while $\sqAB{r}$ gives (r2 ) . Notice how these examples disrupt the normal spacing between lines—a practice to avoid! A userdefined command may have up to nine arguments, numbered 1–9. The simplest examples just allow you to invoke an existing command under a new name. For instance, the \eqref command introduced in Section 7.3 to reference equations (the equation number upright, enclosed in parentheses), would also be useful to reference items for the userdefined list environment enumeratei introduced in Section 15.2.1—see Example 3. Indeed, for the enumeratei environment, we want references to items to be typeset upright, enclosed in parentheses. So if the first item has label First, we could reference it with \eqref{First}, which typesets as (i). But \eqref{First} seems awkward and inappropriate; we are referencing an item not an equation. So define \newcommand{\itemref}[1]{\eqref{#1}} and now we can reference the first item with \itemref{First}, which typesets as (i). Following are some simple examples of userdefined commands with arguments. 1. In the preamble of the source file for this book, I defined \newcommand{\env}[1]{\textnormal{\texttt{#1}}}
15.1
371
Userdefined commands
In this example, the \env command is used to typeset environment names. So the environment name center is typed as \env{center} Again the editorial advantage is obvious. If the editor wants the environment names set in sans serif, only one line in the book has to be changed to alter every occurrence of a typeset environment name: \newcommand{\env}[1]{\textsf{#1}} 2. An argument (e.g., #1) may occur more than once in a definition. A natural example is provided by the \index command (see Section 17.1). Typically, if you wanted to include a phrase, say subdirectly irreducible lattice, in your index, you would have to type this proves that $L$ is a subdirectly irreducible lattice\index{subdirectly irreducible lattice} You could instead define an “index it” command such as \newcommand{\indexit}[1]{#1\index{#1}} The argument of this command is a phrase to be both typeset and included in the index. Using this command, you can type this proves that $L$ is a \indexit{subdirectly irreducible lattice} If you want all such index entries to be typeset in italics, then \indexit should be defined as \newcommand{\indexit}[1]{#1\index{#
[email protected]\textit{#1}}} in which #1 occurs three times. (See Chapter 17 for more information about index commands.) 3. Let us define a command with three arguments for congruences: \newcommand{\congr}[3]{#1\equiv#2\pod{#3}} Now type $\congr{a}{b}{\theta}$ to typeset a ≡ b (θ). In Section 15.1.8, I present another command for typesetting congruences. 4. In the sampart.tex article (see Section 11.3), there are a lot of vectors with only one nonzero entry: i
h. . . , 0, . . . , d, . . . , 0, . . . i the i above the d indicates that it is the ith component of the vector. A command
372
Chapter 15
Customizing LATEX
\vectsup, a vector with a superscript, producing this symbol can be defined as \newcommand{\vectsup}[2]{\langle\dots,0,\dots, \overset{#1}{#2},\dots,0,\dots\rangle} i
\vectsup{i}{d} in a math formula now produces h. . . , 0, . . . , d, . . . , 0, . . . i. 5. Formula 20 of the Formula Gallery (Section 7.9), ϕ · Xn,1 (x + ε2 )2 · · · (x + εn−1 )n−1 (x + εn )n ϕ1 × ε1 ϕ · Xn,1 ϕ · Xn,2 n−1 n · · · (x + εn−1 ) (x + εn ) + I A= n ϕ2 × ε2 ϕ2 × ε1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ϕ · Xn,1 ϕ · Xn,2 ϕ · Xn,n−1 ϕ · Xn,n ··· ϕn × ε1 ϕn × ε2 ϕn × εn−1 ϕn × εn is a good candidate for userdefined commands. By defining \newcommand{\Xquotphi}[2]{% \dfrac{\varphi \cdot X_{n, #1}}% {\varphi_{#2}\times \varepsilon_{#1}}} \newcommand{\exn}[1]{(x+\varepsilon_{#1})^{#1}} the two new commands, \[ \Xquotphi{2}{3} \qquad \exn{n} \] are typeset as ϕ · Xn,2 (x + εn )n ϕ3 × ε2 With these userdefined commands, you can rewrite Formula 20 as follows: \[ \mathbf{A} = \begin{pmatrix} \Xquotphi{1}{1} & \exn{2} & \cdots & \exn{n  1} & \exn{n}\\[10pt] \Xquotphi{1}{2} & \Xquotphi{2}{2} & \cdots & \exn{n  1} &\exn{n}\\ \hdotsfor{5}\\ \Xquotphi{1}{n} & \Xquotphi{2}{n} & \cdots & \Xquotphi{n  1}{n} & \Xquotphi{n}{n} \end{pmatrix} + \mathbf{I}_{n}
15.1
373
Userdefined commands
\] Observe how much shorter this form is than the version shown in the Formula Gallery and how much easier it is to read. It is also easier to reuse in a subsequent article.
15.1.3 Short arguments There are three ways of defining new commands: \newcommand
\renewcommand
\providecommand
We take up the last two in Section 15.1.5. They define commands that can take any number of paragraphs as arguments. The *ed versions of these commands define short commands (see Section 5.3.3) that take a block of text that contains no paragraph break as an argument. For instance, \newcommand{\bigbold}[1]{{\large\bfseries#1}} makes its argument large and bold. So \bigbold{First paragraph. Second paragraph. } prints
First paragraph. Second paragraph. as expected. On the other hand, if you define \newcommand*{\bigbold}[1]{{\large\bfseries#1}} and then attempt to typeset the previous example, you get the error message ! Paragraph ended before \bigbold was complete. \par Short commands are often preferable because of their improved error checking.
374
Chapter 15
15.1.4
Customizing LATEX
Optional arguments You can define a command whose first argument is optional, and provide a default value for this optional argument. To illustrate, let us define the command \newcommand{\SimpleSum}{a_{1}+a_{2}+\dots+a_{n}} $\SimpleSum$ now produces a1 + a2 + · · · + an . Now we change this command so that we can sum from 1 to m if necessary, with n as the default: \newcommand{\BetterSum}[1][n]{a_{1}+a_{2}+\dots+a_{#1}} $\BetterSum$ still produces a1 + a2 + · · · + an , but $\BetterSum[m]$ typesets as a1 + a2 + · · · + am . A \newcommand may have up to nine arguments, but only the first may be optional. The following command has two arguments, one optional: \newcommand{\BestSum}[2][n]{#2_{1}+#2_{2}+\dots+#2_{#1}} Now $\BestSum{a}$ $\BestSum{b}$ $\BestSum[m]{c}$
15.1.5
typesets as typesets as typesets as
a1 + a2 + · · · + an b1 + b2 + · · · + bn c1 + c2 + · · · + cm
Redefining commands LATEX makes sure that you do not inadvertently define a new command with the same name as an existing command (see, for example, page 367). Assuming that you have already defined the \larr command as in Section 15.1.1 (to typeset ←), to redefine \larr, use \renewcommand: \renewcommand{\larr}{\Longleftarrow} and now \larr typesets as ⇐=.
Tip
Use the \renewcommand command sparingly and make sure that you understand the consequences of redefining an existing command. Redefining LATEX commands may cause LATEX to behave in unexpected ways, or even crash. Blind redefinition is the route to madness. See also the discussion in Section 15.7. You can also use \renewcommand to redefine commands defined by LATEX or any package. For instance, the end of proof symbol, \qedsymbol, used by the proof environment, can be changed to the solid black square some people prefer (defined in the amssymb package) with the command
15.1
Userdefined commands
375
\renewcommand{\qedsymbol}{$\blacksquare$} Even better, define \renewcommand{\qedsymbol}{\ensuremath{\blacksquare}} so that you can use \qedsymbol in both text and math mode. Section 15.1.6 has more on redefining names. \renewcommand has a companion, \providecommand. If the command it defines has already been defined, the original command is left unchanged. Otherwise, the \providecommand command acts exactly like \newcommand. For instance, the \bysame command (see Section 10.5.1, page 266) is defined in some document classes as \newcommand{\bysame}{\makebox[3em]{\hrulefill}\thinspace} If you want to use the \bysame command in your bibliography and include this definition in your document, LATEX generates an error message when you typeset your document using a document class that already defines \bysame (all AMS document classes do). However, if you define \bysame in your document using \providecommand: \providecommand{\bysame}% {\makebox[3em]{\hrulefill}\thinspace} the \bysame command typesets correctly whether or not the document class defines it.
15.1.6 Redefining names A number of names, such as Table, List of Tables, Abstract, and so on, are typeset in your document by LATEX. You can easily change these names. For instance, if you are preparing your manuscript for the proceedings of a meeting, and Abstract has to be changed to Summary, you can do so with \renewcommand{\abstractname}{Summary} Table 15.1 lists the commands that define such names in various document classes, along with their default definitions and the major document classes using the commands. It is easy to check whether your document class defines such a command, simply open the appropriate cls file and search for the command. If your document has photographs rather than figures, you could redefine \renewcommand{\figurename}{Photograph} \renewcommand{\listfigurename}{List of Photographs}
376
Chapter 15
Command
Default Value
\abstractname \appendixname \bibname \ccname \chaptername \contentsname \datename \enclname \figurename \headtoname \indexname \keywordsname \listfigurename \listtablename \pagename \partname \proofname \refname \see \seealso \subjclassname
Abstract Appendix Bibliography Cc Chapter Contents Date Enclosure Figure To Index Key words and phrases List of Figures List of Tables Page Part Proof References see see also 1991 Mathematics Subject Classification 2000 Mathematics Subject Classification Table
\subjclassname[2000] \tablename
Customizing LATEX
Defined by Document Class aa, ab, ap, a, p, r aa, ab, ap, a, b, r ab, b, r l ab, b, r aa, ab, ap, a, b, r aa, ab, ap l aa, ab, ap, a, b, r l aa, ab, ap, a, b, r aa, ab, ap aa, ab, ap, a, b, r aa, ab, ap, a, b, r l, p aa, ab, ap, a, b, r aa, ab, ap aa, ap, a aa, ab, ap aa, ab, ap aa, ab, ap aa, ab, ap aa, ab, ap, a, b, r
Document class codes: aa amsart, ab amsbook, ap amsproc, a article, b book, l letter, p proc, and r report Table 15.1: Redefinable name commands in LATEX.
15.1.7
Showing the definitions of commands If you are defining a new command with \newcommand and an error message informs you that the command name is already in use, then it may be useful to find out the existing definition of the command. For instance, the \vects command is defined in sampartu.tex (in the samples folder and in Section 15.4). If you called this new command \vec, you would get the error message ! LaTeX Error: Command \vec already defined.
15.1
Userdefined commands
377
You can find out the definition of the \vec command by getting into interactive mode (see Section D.4) and typing *\show \vec LATEX responds with > \vec=macro: >\mathaccent "017E . \show \vec informing you that \vec is a command, and, specifically, a math accent (see Sections 7.7 and B.8). Now try \hangafter (see Section 5.7.2): *\show \hangafter > \hangafter=\hangafter. \show \hangafter The response indicates that \hangafter is a primitive, defined by TEX itself. Redefining a primitive is not a good idea. Try one more command, \medskip (see Section 5.8.2), to find out how large it is: *\show \medskip > \medskip=macro: >\vspace \medskipamount . The third line indicates that the length is stored in \medskipamount. If we use \show to ask what \medskipamount is defined to be: *\show \medskipamount > \medskipamount=\skip14. we do not get a very useful answer. \medskipamount is unlike most of the commands you have seen so far. It is a length command (see Section 15.5.2), containing the value of \medskip. You can ask for the value of a length command (or parameter) with the \showthe command: *\showthe \medskipamount > 6.0pt plus 2.0pt minus 2.0pt. So \medskip is a vertical space of 6 points that can stretch or shrink by up to 2 points. LATEX has many registers that contain numbers: counters containing integers, such as 3 dimensions such as 10.2pt. For example, \textwidth (see Section 10.6)
378
Chapter 15
Customizing LATEX
lengths, written in the form 6.0pt plus 2.0pt minus 2.0pt, also called a glue or a rubber length (see Sections 15.5.2 and D.3.2) Use the \showthe command to display the value for any of these registers. You can also type the \show and \showthe commands directly into your document rather than go into interactive mode. LATEX’s response appears in the log window, and is saved into the log file.
15.1.8
Delimited commands You can define new commands in TEX using characters and symbols to delimit arguments. Such delimited commands provide a way to write more readable source documents. First we have to learn how to define a command using TEX’s \def command. Type \def, followed by the new command name (not in braces), then the definition in braces. For example, the first command defined in Section 15.1.1, \newcommand{\larr}{\leftarrow} could be typed \def\larr{\leftarrow} TEX’s \def command does not check whether a new command name is already in use, so \def behaves differently from the LATEX’s \newcommand, \renewcommand, and \providecommand (see Section 15.1.5). If the \larr command was defined previously, the original definition is overwritten.
Tip
It is your responsibility to ensure that your command name is unique when you define a command using \def. LATEX provides no protection. Use the techniques introduced in Section 15.1.7 to check a name before you define a command with \def. Now we can start discussing delimited commands with a simple example, defining a command for vectors: \def\vect{\langle#1\rangle} Note that \vect is a command with one argument, #1. When invoked, it typesets h, the argument, and then i. In the definition of \vect, the argument #1 is delimited by < and >. When the command is invoked, the argument must be delimited the same way. So to typeset the vector ha, bi, we invoke \vect with \vect
15.1
379
Userdefined commands
This looks somewhat like a vector, and the name \vect serves as a reminder. You have to be careful with delimited commands because the math spacing rules (see Section 7.2) do not hold in either the definition or the invocation. So if there is a space before #1, in the definition of \vect, \def\vect< #1>{\langle#1\rangle} then $\vect$ results in the error message ! Use of \vect doesn’t match its definition. l.12 $\vect$ which is clear enough. If the space is on the other side of the #1, as in \def\vect{\langle#1\rangle} the error message is slightly more confusing: Runaway argument? a,b>$ ! Paragraph ended before \vect was complete. \par The moral is that if you use delimited commands, you must be very careful that each invocation exactly matches the definition. In Example 3 of Section 15.1.2, we introduced a command with three arguments for typing congruences: \newcommand{\congr}[3]{#1\equiv#2\pod{#3}} $\congr{a}{b}{\theta}$ produces a ≡ b (θ). This command is easy to remember, but it does not make the source file more readable. For that, we use a delimited command. Let us redo the congruence example with a delimited command \def\congr#1=#2(#3){#1\equiv#2\pod{#3}} so that $\congr a=b(\theta)$ produces a ≡ b (θ). In the source document, the formula \congr a=b(\theta) looks a bit like the typeset congruence and it is easier to read. I included this definition in the newlattice.sty command file (see Section 15.3). There is only one catch. Suppose you want to typeset the formula x = a ≡ b (θ) If you type $\congr x=a=b(\theta)$, LATEX typesets it as x ≡ a = b (θ). Indeed, x is delimited on the right by the first =, so LATEX believes that the first argument is x.
380
Chapter 15
Customizing LATEX
The second argument is delimited by the first = and the left parenthesis, so it is a=b. In such cases, you can help LATEX find the correct first argument by enclosing it in braces: $\congr{x=a}=b(\theta)$ Here is our final example. In Section 5.3.1 we discuss the problem of typing a command such as \TeX (the example there was \today) in the form \TeX\ so that it is typeset as a separate word. The problem is that if you type \TeX without the trailing \ , TEX is merged with the next word, and there is no error message to warn you. One solution is to use a delimited command: \def\tex/{\TeX} Now to get TEX, type \tex/. If a space is needed after it, type \tex/ . If you forget the closing /, you get an error message. A better solution to this problem is the use of the xspace package—provided you do not want to typeset something like TEXbook (see Section 15.1.1). However, many documents use the delimited construct (including the AMS documentation), so you should be familiar with it.
15.2
Userdefined environments Most userdefined commands are new commands. Userdefined environments, as a rule, are built on existing environments. We start with such userdefined environments (Section 15.2.1) and then proceed to investigate arguments (Section 15.2.2) optional arguments (Section 15.2.3) short arguments (Section 15.2.4) Finally, we discuss how to define brandnew environments (Section 15.2.5).
15.2.1
Modifying existing environments If you do not like the name of the proof environment and would prefer to use the name demo, define \newenvironment{demo} {\begin{proof}} {\end{proof}} Note that this does not change how the environment is typeset, only the way it is invoked.
15.2
Userdefined environments
381
To modify an existing environment, oldenv, type \newenvironment{name } {begin_text } {end_text } where begin text contains the command \begin{oldenv } and end text contains the command \end{oldenv }.
Tip
Do not give a new environment the name of an existing command or environment. For instance, if you define \newenvironment{parbox} {...} {...} you get the error message ! LaTeX Error: Command \parbox already defined. If there is an error in such a userdefined environment, the message generated refers to the environment that was modified, not to your environment. For instance, if you misspell proof as prof when you define \newenvironment{demo} {\begin{prof}} {\end{proof}} then at the first use of the demo environment you get the message ! LaTeX Error: Environment prof undefined. l.13 \begin{demo} If you define \newenvironment{demo} {\begin{proof}\em} {\end{prof}} at the first use of demo you get the message ! LaTeX Error: \begin{proof} on input line 5 ended by \end{prof}. l.14 \end{demo}
382
Chapter 15
Customizing LATEX
Here are four more examples of modified environments. 1. The command \newenvironment{demo} {\begin{proof}\em} {\end{proof}} defines a demo environment that typesets an emphasized proof. Note that the scope of \em is the demo environment. 2. The following example defines a very useful environment. It takes an argument to be typeset as the name of a theorem: \newtheorem*{namedtheorem}{\theoremname} \newcommand{\theoremname}{testing} \newenvironment{named}[1]{ \renewcommand{\theoremname}{#1} \begin{namedtheorem}} {\end{namedtheorem}} For example, \begin{named}{Name of the theorem} Body of theorem. \end{named} produces Name of the theorem. Body of theorem. in the style appropriate for the \newtheorem* declaration. This type of environment is often used to produce an unnumbered Main Theorem (see Section 15.4) or when typesetting an article or book in which the theorem numbering is already fixed, for instance, when publishing a book in LATEX that was originally typeset by another typesetting system. 3. In Sections 6.2.4 and 12.3.1, we came across the enumerate package, which allows you to customize the enumerate environment. If the enumerate package is loaded, you can invoke the enumerate environment with an optional argument specifying how the counter should be typeset, for instance, with the option [\upshape (i)], \begin{enumerate}[\upshape (i)] \item First item\label{First} \end{enumerate} items are numbered (i), (ii), and so on.
15.2
Userdefined environments
383
So now we define \newenvironment{enumeratei}{\begin{enumerate}% [\upshape (i)]}% {\end{enumerate}} and we can invoke the new environment with (see Sections 15.3 and 15.4) \begin{enumeratei} \item \label{ } \end{enumeratei} Reference items in the enumeratei environment with the \itemref command introduced in Section 15.1.2. 4. If you want to define an environment for displaying text that is numbered as an equation, you might try \newenvironment{texteqn} {\begin{equation} \begin{minipage}{0.9\linewidth}} {\end{minipage} \end{equation}} But there is a problem. If you use this environment in the middle of a paragraph, an interword space appears at the beginning of the first line after the environment. To remove this unwanted space, use the \ignorespacesafterend command, as in \newenvironment{texteqn} {\begin{equation} \begin{minipage}{0.9\linewidth}} {\end{minipage} \end{equation} \ignorespacesafterend} Examples 2 and 3 are included in the newlattice.sty command file (see Section 15.3). See the sample article, sampartu.tex in Section 15.4, for some instances of their use. See Section 15.6.3 for custom lists as userdefined environments. Redefine an existing environment with the \renewenvironment command. It is similar to the \renewcommand command (see Section 15.1.5). There are some environments you cannot redefine; for instance, verbatim and all the AMS multiline math environments.
15.2.2 Arguments An environment defined by the \newenvironment command can take arguments (see Example 2 in Section 15.2.1), but they can only be used in the begin text argument of the \newenvironment command. Here is a simple example. Define a theorem proclamation in the preamble (see Section 6.4), and then define a theorem that can be referenced:
384
Chapter 15
Customizing LATEX
\newenvironment{theoremRef}[1] {\begin{theorem}\label{T:#1}} {\end{theorem}} This is invoked with \begin{theoremRef}{label } The theoremRef environment is a modified environment. It is a theorem that can be referenced (with the \ref and \pageref commands, of course) and it invokes the theorem environment when it defines T:label to be the label for crossreferencing.
15.2.3
Optional arguments with default values The first argument of an environment created with the \newenvironment command may be an optional argument with a default value. For example, \newenvironment{narrow}[1][3in] {\noindent\begin{minipage}{#1}} {\end{minipage}} creates a narrow environment. By default, it sets the body of the environment in a 3inch wide box, with no indentation. So \begin{narrow} This text was typeset in a \texttt{narrow} environment, in a 3inch wide box, with no indentation. \end{narrow} typesets as This text was typeset in a narrow environment, in a 3inch wide box, with no indentation. You can also give an optional argument to specify the width. For example, \begin{narrow}[3.5in] This text was typeset in a \texttt{narrow} environment, in a 3inch wide box, with no indentation. \end{narrow} which produces the following false statement: This text was typeset in a narrow environment, in a 3inch wide box, with no indentation.
15.2
Userdefined environments
385
15.2.4 Short contents We have discussed two commands that define new environments, \newenvironment and \renewenvironment These commands allow you to define environments whose contents (begin text and end text ; see page 381) can include any number of paragraphs. The *ed versions of these commands define short environments whose contents cannot contain a paragraph break (a blank line or a \par command).
15.2.5 Brandnew environments Some userdefined environments are not modifications of existing environments. Here are two examples: 1. A command remains effective only within its scope (see Section 5.3.2). Now suppose that you want to make a change, say redefining a counter, for only a few paragraphs. You could simply place braces around these paragraphs, but they are hard to see. So define \newenvironment{exception} {\relax} {\relax} and then \begin{exception} new commands body \end{exception} The environment stands out better than a pair of braces, reminding you later about the special circumstances. The \relax command does nothing, but it is customary to include a \relax command in such a definition to make it more readable. 2. In this example, we define a new environment that centers its body vertically on a new page: \newenvironment{vcenterpage} {\newpage\vspace*{\fill}} {\vspace*{\fill}\par\pagebreak} For \vspace, see Section 5.8.2 and for \fill, see the last subsection in Section 15.5.
386
Chapter 15
15.3
Customizing LATEX
A custom command file Userdefined commands, of course, are a matter of individual need and taste. I have collected some commands for writing papers in lattice theory in the newlattice.sty file, which you can find in the samples folder (see page 4). I hope that this model helps you to develop a command file of your own. Please remember that everything we discuss in this section is a reflection of my work habits. Many experts disagree with one or another aspect of the way I define the commands, so take whatever suits your needs. And keep in mind the dangers of customization discussed in Section 15.7.
Tip
Some journals do not permit the submission of a separate custom command file. For such journals, just copy the needed userdefined commands into the preamble of the article. This file is named newlattice.sty. It can be loaded with \usepackage. This has a number of advantages. Your command names should be mnemonic. If you cannot easily remember a command’s name, rename it. The implication here is that your command file should not be very large unless you have an unusual ability to recall abbreviations. Here are the first few lines of the newlattice.sty command file: % newlattice.sty % New command file for lattice papers \NeedsTeXFormat{LaTeX2e}[2005/12/01] \ProvidesPackage{newlattice}[2006/03/15 v1.2 New commands for lattices] \RequirePackage{amsmath} \RequirePackage{amssymb} \RequirePackage{latexsym} \RequirePackage[mathscr]{eucal} \RequirePackage{verbatim} \RequirePackage{enumerate} \RequirePackage{xspace} The line \NeedsTeXFormat{LaTeX2e}[2005/12/01] gives an error message if a document loading the newlattice package is typeset with LATEX 2.09 or with an older version of the standard LATEX. The next line provides information that is written in your log file. The next seven lines declare what packages are required. If some of these packages have not yet been loaded, then the missing packages are loaded. A package loaded
15.3
A custom command file
387
with \RequirePackage is not read in again. Being able to specify the packages we need is one of the great advantages of command files. When I write a document, the packages are there if I need them. You may want some justification for the inclusion of two of these packages in this list. The verbatim package is on the list so that I can use the comment environment to comment out large blocks of text (see Section 5.5.1), which is useful for finding errors and typesetting only parts of a longer document—but do not forget to remove your comments before submission. The enumerate package is on the list because the enumeratei and enumeratea environments, defined in newlattice.sty, require it. If you start your article with \documentclass{amsart} \usepackage{newlattice} then the \listfiles command (see Section D.3.4) produces the following list when your document is typeset: *File List* amsart.cls 2004/08/06 v2.20 amsmath.sty 2000/07/18 v2.13 AMS math features amstext.sty 2000/06/29 v2.01 amsgen.sty 1999/11/30 v2.0 amsbsy.sty 1999/11/29 v1.2d amsopn.sty 1999/12/14 v2.01 operator names umsa.fd 2002/01/19 v2.2g AMS font definitions amsfonts.sty 2001/10/25 v2.2f newlattice.sty 2006/03/15 New commands for lattices v1.2 amssymb.sty 2002/01/22 v2.2d enumerate.sty 1999/03/05 v3.00 enumerate extensions (DPC) graphicx.sty 1999/02/16 v1.0f Enhanced LaTeX Graphics (DPC,SPQR) keyval.sty 1999/03/16 v1.13 key=value parser (DPC) trig.sty 1999/03/16 v1.09 sin cos tan (DPC) graphics.cfg 2005/02/03 v1.3 graphics configuration of teTeX/TeXLive pdftex.def 2002/06/19 v0.03k graphics/color for pdftex umsa.fd 2002/01/19 v2.2g AMS font definitions umsb.fd 2002/01/19 v2.2g AMS font definitions ueuf.fd 2002/01/19 v2.2g AMS font definitions ***********
388
Chapter 15
Customizing LATEX
Now we continue with newlattice.sty. After the introductory section dealing with LATEX and the packages, we define some commands for writing about lattices and sets: % Lattice operations \newcommand{\jj}{\vee}% join \newcommand{\mm}{\wedge}% meet \newcommand{\JJ}{\bigvee}% big join \newcommand{\MM}{\bigwedge}% big meet \newcommand{\JJm}[2]{\JJ(\,#1\mid#2\,)}% big join with a middle \newcommand{\MMm}[2]{\MM(\,#1\mid#2\,)}% big meet with a middle % Set operations \newcommand{\uu}{\cup}% union \newcommand{\ii}{\cap}% intersection \newcommand{\UU}{\bigcup}% big union \newcommand{\II}{\bigcap}% big intersection \newcommand{\UUm}[2]{\UU(\,#1\mid#2\,)}% big union with a middle \newcommand{\IIm}[2]{\II(\,#1\mid#2\,)} % big intersection with a middle % Sets \newcommand{\contd}{\subseteq}% contained in \newcommand{\ncontd}{\nsubseteq}% not \contd \newcommand{\scontd}{\subset}% strictly contained in \newcommand{\contg}{\supseteq}% containing with equality \newcommand{\ncontg}{\nsupseteq}% not \contg \newcommand{\nin}{\notin}% not \in \newcommand{\empset}{\varnothing}% the empty set \newcommand{\set}[1]{\{#1\}}% set \newcommand{\setm}[2]{\{\,#1\mid#2\,\}}% set with a middle \def\vect{\langle#1\rangle}% vector So $a \jj b$ produces a ∨ b and $A \contd B$ produces A ⊆ B, and so on. The original commands are not redefined, so if a coauthor prefers $a \vee b$ to $a \jj b$, the \vee command is available. The commands with a “middle” are exemplified by \setm: $\setm{x \in R}{x^2 \leq 2}$ typesets as { x ∈ R  x2 ≤ 2 }. Using the \set command, we can type the set {a, b} as $\set{a,b}$, which is easier to read than $\{a,b\}$. Similarly, we type $\vect$ for the vector ha, bi,
15.3
A custom command file
389
so it looks like a vector. Next in newlattice.sty I map the Greek letters to easy to remember commands. For some, I prefer to use the variants, but that is a matter of individual taste. It is also a matter of taste whether or not to change the commands for the Greek letters at all, and how far one should go in changing commonly used commands. % Greek letters \newcommand{\Gra}{\alpha} \newcommand{\Grb}{\beta} \newcommand{\Grc}{\chi} \newcommand{\Grd}{\delta} \renewcommand{\Gre}{\varepsilon} \newcommand{\Grf}{\varphi} \renewcommand{\Grg}{\gamma} \newcommand{\Grh}{\eta} \newcommand{\Gri}{\iota} \newcommand{\Grk}{\kappa} \newcommand{\Grl}{\lambda} \newcommand{\Grm}{\mu} \newcommand{\Grn}{\nu} \newcommand{\Gro}{\omega} \newcommand{\Grp}{\pi} \newcommand{\Grq}{\theta} \newcommand{\Grr}{\varrho} \newcommand{\Grs}{\sigma} \newcommand{\Grt}{\tau} \newcommand{\Gru}{\upsilon} \newcommand{\Grv}{\vartheta} \newcommand{\Grx}{\xi} \newcommand{\Gry}{\psi} \newcommand{\Grz}{\zeta} \newcommand{\GrG}{\Gamma} \newcommand{\GrD}{\Delta} \newcommand{\GrF}{\Phi} \newcommand{\GrL}{\Lambda} \newcommand{\GrO}{\Omega} \newcommand{\GrP}{\Pi} \newcommand{\GrQ}{\Theta} \newcommand{\GrS}{\Sigma} \newcommand{\GrU}{\Upsilon} \newcommand{\GrX}{\Xi} \newcommand{\GrY}{\Psi}
390
Chapter 15
Customizing LATEX
I also introduce some new names for text font commands by abbreviating text to t (so that \textbf becomes \tbf) and for math font commands by abbreviating math to m (so that \mathbf becomes \mbf). % Font commands \newcommand{\tbf}{\textbf}% text bold \newcommand{\tit}{\textit}% text italic \newcommand{\tsl}{\textsl}% text slanted \newcommand{\tsc}{\textsc}% text small cap \newcommand{\ttt}{\texttt}% text typewriter \newcommand{\trm}{\textrm}% text roman \newcommand{\tsf}{\textsf}% text sans serif \newcommand{\tup}{\textup}% text upright \newcommand{\mbf}{\mathbf}% math bold \providecommand{\mit}{\mathit}% math italic \newcommand{\msf}{\mathsf}% math sans serif \newcommand{\mrm}{\mathrm}% math roman \newcommand{\mtt}{\mathtt}% math typewriter The math alphabets are invoked as commands with arguments: \Bold for bold, \Cal for calligraphic, \DD for blackboard bold (double), and \Frak for fraktur (German Gothic) (see Section 8.3.2). Notice that \Cal and \Euler are different because of the option mathscr of the eucal package (see Section 8.3.1). \newcommand{\Bold}[1]{\boldsymbol{#1}} % Bold math symbol, use as \Bold{\alpha} \newcommand{\Cal}[1]{\mathcal{#1}} % Calligraphics  only caps, use as \Cal{A} \newcommand{\DD}[1]{\mathbb{#1}} % Doubled  blackboard bold  only caps, use as \DD{A} \newcommand{\Euler}[1]{\mathscr{#1}} % Euler Script  only caps, use as \Euler{A} \newcommand{\Frak}[1]{\mathfrak{#1}}% Fraktur, use as \Frak{a} \Bold{A} typesets bold italic A, A; to get upright bold, use \mathbf{A}, which typesets as A. Here are some commands of importance in lattice theory: % Constructs \DeclareMathOperator{\Id}{Id} \DeclareMathOperator{\Fil}{Fil} \DeclareMathOperator{\Con}{Con} \DeclareMathOperator{\Aut}{Aut} \DeclareMathOperator{\Sub}{Sub}
15.3
A custom command file
391
\DeclareMathOperator{\Pow}{Pow} \DeclareMathOperator{\Part}{Part} \DeclareMathOperator{\Ker}{Ker} \newcommand{\Ji}[1]{\tup{J}(#1)} %join irreducible \newcommand{\Mi}[1]{\tup{M}(#1)} %meet irreducible % Generated by \newcommand{\con}[1]{\tup{con}(#1)} \newcommand{\sub}[1]{\tup{sub}(#1)} \newcommand{\id}[1]{\tup{id}(#1)} \newcommand{\fil}[1]{\tup{f{}il}(#1)} \newcommand{\Downg}{\downarrow\!}% downset generated by Here are a few more commands and environments: % Miscellaneous \newcommand{\newl}{\newline} \newcommand{\overl}[1]{\overline{#1}} \newcommand{\underl}[1]{\underline{#1}} \providecommand{\bysame}{\makebox[3em]% {\hrulefill}\thinspace} \newcommand{\iso}{\cong}% isomorphic \def\congr#1=#2(#3){#1\equiv#2\pod{#3}} %congruence, use it as \congr a=b(\theta)% \newcommand{\itemref}[1]{\eqref{#1}} \newenvironment{enumeratei}{\begin{enumerate}% [\upshape (i)]}{\end{enumerate}} %produces (i), (ii), etc. Reference with \itemref \newenvironment{enumeratea}{\begin{enumerate}% [\upshape (a)]}{\end{enumerate}} %produces (a), (b), etc. Reference with \itemref \theoremstyle{plain} \newtheorem*{namedtheorem}{\theoremname} \newcommand{\theoremname}{testing} \newenvironment{named}[1]{\renewcommand{\theoremname}{#1} \begin{namedtheorem}} {\end{namedtheorem}} %use it as \begin{named}{Name of theorem} %Body of theorem \end{named} \endinput For the \congr command see Section 15.1.8. The enumeratei and named environments are discussed in Section 15.2.1. The enumeratea environment is similar.
392
Chapter 15
Customizing LATEX
This command file, like all command files, is terminated with the \endinput command. In Section 18.3.2, we discuss the same rule for files that are \included. My newlattice.sty evolves with time. I keep a copy in the folder of every article I write. This way, even years later, with the command file much changed, I can typeset the article with no problem. The \TextOrMath command (see Section 12.3) is very useful for command files. For instance, we can use it to define our Greek letters, such as \newcommand{\Gra}{\TextOrMath{$\alpha$}{\alpha}} Then we can use \Gra both in text and math to produce α. Read the discussion in Sections A.1.6 and A.2.6 where to put your custom command file.
15.4
The sample article with userdefined commands In this section, we look at the sampartu.tex sample article (also in the samples folder), which is a rewrite of the sampart.tex sample article (see Section 11.3 and the samples folder) utilizing the userdefined commands collected in the command file newlattice.sty (see Section 15.3 and the samples folder). % Sample file: sampartu.tex % The sample article % with userdefined commands and environments \documentclass{amsart} \usepackage{newlattice} \theoremstyle{plain} \newtheorem{theorem}{Theorem} \newtheorem{corollary}{Corollary} \newtheorem{lemma}{Lemma} \newtheorem{proposition}{Proposition} \theoremstyle{definition} \newtheorem{definition}{Definition} \theoremstyle{remark} \newtheorem*{notation}{Notation} \numberwithin{equation}{section} \newcommand{\Prodm}[2]{\GrP(\,#1\mid#2\,)} % product with a middle
15.4
The sample article with userdefined commands \newcommand{\Prodsm}[2]{\GrP^{*}(\,#1\mid#2\,)} % product * with a middle \newcommand{\vectsup}[2]{\vect}% special vector \newcommand{\Dsq}{D^{\langle2\rangle}} \begin{document} \title[Completesimple distributive lattices] {A construction of completesimple\\ distributive lattices} \author{George~A. Menuhin} \address{Computer Science Department\\ University of Winnebago\\ Winnebago, Minnesota 23714} \email{
[email protected]} \urladdr{http://math.uwinnebago.edu/homepages/menuhin/} \thanks{Research supported by the NSF under grant number~23466.} \keywords{Complete lattice, distributive lattice, complete congruence, congruence lattice} \subjclass[2000]{Primary: 06B10; Secondary: 06D05} \date{March 15, 2006} \begin{abstract} In this note we prove that there exist \emph{completesimple distributive lattices,} that is, complete distributive lattices in which there are only two complete congruences. \end{abstract} \maketitle \section{Introduction}\label{S:intro} In this note we prove the following result: \begin{named}{Main Theorem} There exists an infinite complete distributive lattice $K$ with only the two trivial complete congruence relations. \end{named} \section{The $\Dsq$ construction}\label{S:Ds} For the basic notation in lattice theory and universal algebra, see Ferenc~R. Richardson~\cite{fR82} and George~A. Menuhin~\cite{gM68}.
393
394
Chapter 15
Customizing LATEX
We start with some definitions: \begin{definition}\label{D:prime} Let $V$ be a complete lattice, and let $\Frak{p} = [u, v]$ be an interval of $V$. Then $\Frak{p}$ is called \emph{completeprime} if the following three conditions are satisfied: \begin{enumeratei} \item $u$ is meetirreducible but $u$ is \emph{not} completely meetirreducible;\label{mi} \item $v$ is joinirreducible but $v$ is \emph{not} completely joinirreducible;\label{ji} \item $[u, v]$ is a completesimple lattice.\label{cs} \end{enumeratei} \end{definition} Now we prove the following result: \begin{lemma}\label{L:Dsq} Let $D$ be a complete distributive lattice satisfying conditions \itemref{mi} and~\itemref{ji}. Then $\Dsq$ is a sublattice of $D^{2}$; hence $\Dsq$ is a lattice, and $\Dsq$ is a complete distributive lattice satisfying conditions \itemref{mi} and~\itemref{ji}. \end{lemma} \begin{proof} By conditions~\itemref{mi} and \itemref{ji}, $\Dsq$ is a sublattice of $D^{2}$. Hence, $\Dsq$ is a lattice. Since $\Dsq$ is a sublattice of a distributive lattice, $\Dsq$ is a distributive lattice. Using the characterization of standard ideals in Ernest~T. Moynahan~\cite{eM57}, $\Dsq$ has a zero and a unit element, namely, $\vect$ and $\vect$. To show that $\Dsq$ is complete, let $\empset \ne A \contd \Dsq$, and let $a = \JJ A$ in $D^{2}$. If $a \in \Dsq$, then $a = \JJ A$ in $\Dsq$; otherwise, $a$ is of the form $\vect$ for some $b \in D$ with $b < 1$. Now $\JJ A = \vect$ in $D^{2}$, and
15.4
The sample article with userdefined commands the dual argument shows that $\MM A$ also exists in $D^{2}$. Hence $D$ is complete. Conditions \itemref{mi} and~\itemref{ji} are obvious for $\Dsq$. \end{proof} \begin{corollary}\label{C:prime} If $D$ is completeprime, then so is $\Dsq$. \end{corollary} The motivation for the following result comes from SooKey Foo~\cite{sF90}. \begin{lemma}\label{L:ccr} Let $\GrQ$ be a complete congruence relation of $\Dsq$ such that \begin{equation}\label{E:rigid} \congr \vect=\vect(\GrQ), \end{equation} for some $d \in D$ with $d < 1$. Then $\GrQ = \Gri$. \end{lemma} \begin{proof} Let $\GrQ$ be a complete congruence relation of $\Dsq$ satisfying \itemref{E:rigid}. Then $\GrQ = \Gri$. \end{proof} \section{The $\Grp^{*}$ construction}\label{S:P*} The following construction is crucial to our proof of the Main~Theorem: \begin{definition}\label{D:P*} Let $D_{i}$, for $i \in I$, be complete distributive lattices satisfying condition~\itemref{ji}. Their $\Grp^{*}$ product is defined as follows: \[ \Prodsm{ D_{i} }{i \in I} = \Prodm{ D_{i}^{} }{i \in I} + 1; \] that is, $\Prodsm{ D_{i} }{i \in I}$ is $\Prodm{D_{i}^{}}{i \in I}$ with a new unit element. \end{definition}
395
396
Chapter 15
Customizing LATEX
\begin{notation} If $i \in I$ and $d \in D_{i}^{}$, then \[ \vectsup{i}{d} \] is the element of $\Prodsm{ D_{i} }{i \in I}$ whose $i$th component is $d$ and all the other components are $0$. \end{notation} See also Ernest~T. Moynahan~\cite{eM57a}. Next we verify: \begin{theorem}\label{T:P*} Let $D_{i}$, for $i \in I$, be complete distributive lattices satisfying condition~\itemref{ji}. Let $\GrQ$ be a complete congruence relation on $\Prodsm{ D_{i} }{i \in I}$. If there exist $i \in I$ and $d \in D_{i}$ with $d < 1_{i}$ such that for all $d \leq c < 1_{i}$, \begin{equation}\label{E:cong1} \congr\vectsup{i}{d}=\vectsup{i}{c}(\GrQ), \end{equation} then $\GrQ = \Gri$. \end{theorem} \begin{proof} Since \begin{equation}\label{E:cong2} \congr\vectsup{i}{d}=\vectsup{i}{c}(\GrQ), \end{equation} and $\GrQ$ is a complete congruence relation, it follows from condition~\itemref{cs} that \begin{equation}\label{E:cong} \begin{split} &\langle \dots, \overset{i}{d}, \dots, 0, \dots \rangle\\ &\equiv \bigvee ( \langle \dots, 0, \dots, \overset{i}{c},\dots, 0,\dots \rangle \mid d \leq c < 1) \equiv 1 \pmod{\Theta}. \end{split} \end{equation}
15.4
The sample article with userdefined commands Let $j \in I$, for $j \neq i$, and let $a\in D_{j}^{}$. Meeting both sides of the congruence \itemref{E:cong} with $\vectsup{j}{a}$, we obtain \begin{equation}\label{E:comp} \begin{split} 0 &= \vectsup{i}{d} \mm \vectsup{j}{a}\\ &\equiv \vectsup{j}{a}\pod{\GrQ}. \end{split} \end{equation} Using the completeness of $\GrQ$ and \itemref{E:comp}, we get: \begin{equation}\label{E:cong3} \congr{0=\JJm{\vectsup{j}{a}}{ a \in D_{j}^{} }} ={1}(\GrQ), \end{equation} hence $\GrQ = \Gri$. \end{proof} \begin{theorem}\label{T:P*a} Let $D_{i}$, for $i \in I$, be complete distributive lattices satisfying conditions \itemref{ji} and~\itemref{cs}. Then $\Prodsm{ D_{i} }{i \in I}$ also satisfies conditions~\itemref{ji} and \itemref{cs}. \end{theorem} \begin{proof} Let $\GrQ$ be a complete congruence on $\Prodsm{ D_{i} }{i \in I}$. Let $i \in I$. Define \begin{equation}\label{E:dihat} \widehat{D}_{i} = \setm{ \vectsup{i}{d} } { d \in D_{i}^{} } \uu \set{1}. \end{equation} Then $\widehat{D}_{i}$ is a complete sublattice of $\Prodsm{ D_{i} }{i \in I}$, and $\widehat{D}_{i}$ is isomorphic to $D_{i}$. Let $\GrQ_{i}$ be the restriction of $\GrQ$ to $\widehat{D}_{i}$. Since $D_{i}$ is completesimple, so is $\widehat{D}_{i}$, hence $\GrQ_{i}$ is $\Gro$ or $\Gri$. If $\GrQ_{i}=\Gro$, for all $i\in I$, then $\GrQ = \Gro$. If there is an $i \in I$, such that $\GrQ_{i} = \Gri$, then $\congr0=1(\GrQ)$, and hence $\GrQ = \Gri$. \end{proof}
397
398
Chapter 15
Customizing LATEX
The Main Theorem follows easily from Theorems~\ref{T:P*} and \ref{T:P*a}. \begin{thebibliography}{9} \bibitem{sF90} SooKey Foo, \emph{Lattice Constructions}, Ph.D. thesis, University of Winnebago, Winnebago, MN, December, 1990. \bibitem{gM68} George~A. Menuhin, \emph{Universal algebra}. D.~Van Nostrand, Princeton, 1968. \bibitem{eM57} Ernest~T. Moynahan, \emph{On a problem of M. Stone}, Acta Math. Acad. Sci. Hungar. \tbf{8} (1957), 455460. \bibitem{eM57a} \bysame, \emph{Ideals and congruence relations in lattices}.~II, Magyar Tud. Akad. Mat. Fiz. Oszt. K\"{o}zl. \tbf{9} (1957), 417434 (Hungarian). \bibitem{fR82} Ferenc~R. Richardson, \emph{General Lattice Theory}. Mir, Moscow, expanded and revised ed., 1982 (Russian). \end{thebibliography} \end{document}
15.5
Numbering and measuring LATEX stores integers in counters. For example, the section counter contains the current section number. Distance measurements are saved in length commands. For instance, the \textwidth command contains the width of the text. For this book, \textwidth is set to 345.0 points. In this section, we take a closer look at counters and length commands.
15.5
399
Numbering and measuring
15.5.1 Counters Counters may be defined by LATEX, by document classes, by packages, or by the user. Standard LATEX counters LATEX automatically generates numbers for equations, sections, theorems, and so on. Each such number is stored in a counter. Table 15.2 shows the standard LATEX counters. Their names are more or less selfexplanatory. In addition, for every proclamation name, there is a matching counter called name (see Section 6.4). Setting counters The command for setting a counter’s value is \setcounter. When LATEX generates a number, it first increments the appropriate counter, so if you want the next chapter to be numbered 3, you should set the chapter counter to 2 by typing \setcounter{chapter}{2} before the \chapter command. The only exception to this rule is the page number, which is first used to number the current page, and then incremented. If you wanted to set the current page number to 63, you would include the command \setcounter{page}{63} somewhere in the page. LATEX initializes and increments its standard counters automatically. Sometimes you may want to manipulate them yourself. To typeset only chapter3.tex, the third chapter of your book, start with \setcounter{chapter}{2} \include{chapter3} and when chapter3.tex is typeset, the chapter is properly numbered. You can also type \setcounter{page}{63} equation figure footnote mpfootnote page table
part chapter section subsection subsubsection paragraph subparagraph
enumi enumii enumiii enumiv
Table 15.2: Standard LATEX counters.
400
Chapter 15
Customizing LATEX
if the first page of this chapter is supposed to be 63. Of course, the preferred way to typeset parts of a larger document is with the \includeonly command (see Section 18.3.2).
Tip
If you need to manipulate counters, always look for solutions in which LATEX does the work for you.
Defining new counters You can define your own counters. For example, \newcounter{mycounter} makes mycounter a new counter. In the definition, you can use an optional argument, the name of another counter: \newcounter{mycounter}[basecounter] which automatically resets mycounter to 0 if basecounter changes value. This command has the same form as the command LATEX uses internally for tasks such as numbering theorems and subsections within sections.
Rule
New counters New counters should be defined in the preamble of the document. They should not be defined in a file read in with an \include command (see Section 18.3.2). Let us suppose that you define a new counter, mycounter, in chapter5.tex, which is made part of your whole document with an \include command. When you typeset your document with \includeonly commands not including chapter5.tex, you get an error message, such as ! LaTeX Error: No counter ’mycounter’ defined. Counter styles The value of counter can be displayed in the typeset document with the command \thecounter If you want to change the counter’s appearance when typeset, issue the command \renewcommand{\thecounter }{new_style }
15.5
401
Numbering and measuring Style
Command
Sample
Arabic Lowercase Roman Uppercase Roman Lowercase Letters Uppercase Letters
\arabic{counter } \roman{counter } \Roman{counter } \alph{counter } \Alph{counter }
1, 2, . . . i, ii, . . . I, II, . . . a, b, . . . , z A, B, . . . , Z
Table 15.3: Counter styles.
where new style specifies the counter modified as shown in Table 15.3. The default style is arabic. For instance, if you give the command \renewcommand{\thetheorem}{\Alph{theorem}} then the theorems appear as Theorem A, Theorem B, . . . Here is a more complicated example for a book: \renewcommand{\thechapter}{\arabic{chapter}} \renewcommand{\thesection}{\thechapter\arabic{section}} \renewcommand{\thesubsection}% {\thechapter\arabic{section}.\arabic{subsection}} With these definitions, Section 1 of Chapter 3 is numbered in the form 31 and Subsection 2 of Section 1 of Chapter 3 is numbered in the form 31.2. The \pagenumbering command is a shorthand method for setting the page numbering in a given style. For instance, \pagenumbering{roman} numbers pages as i, ii, and so on. The subequations environment (see Section 8.5) uses parentequation as the counter for the whole equation group and it uses equation as the counter for the subequations. To change the default format of the equation numbers from (2a), (2b), and so forth, to (2i), (2ii), and so on, type the following line inside the subequations environment \renewcommand{\theequation}% {\theparentequation\roman{equation}} If you want equation numbers like (2.i), (2.ii), and so on, type \renewcommand{\theequation}% {\theparentequation.\roman{equation}}
402
Chapter 15
Customizing LATEX
Counter arithmetic The \stepcounter{counter} command increments counter and sets all the counters that were defined with the optional argument counter to 0. The variant \refstepcounter{counter} does the same, and also sets the value for the next \label command. You can do some arithmetic with the command \addtocounter{counter }{n } where n is an integer. For example, \setcounter{counter}{5} \addtocounter{counter}{2} sets counter to 7. The value stored in a counter can be accessed using the \value command, which is mostly used with the \setcounter or \addtocounter commands. For instance, you can set counter to equal the value of another counter, oldcounter, by typing \setcounter{counter}{\value{oldcounter}} Here is a typical example of counter manipulation. You might want a theorem (invoked in a theorem environment) to be followed by several corollaries (each in a corollary environment) starting with Corollary 1. In other words, Theorem 1 should be followed by Corollary 1, Corollary 2, and so forth and so should Theorem 3. By default, LATEX numbers the next corollary as Corollary 3, even if it follows another theorem. To tell LATEX to start numbering the corollaries from 1 again, issue the command \setcounter{corollary}{0} after each theorem. But such a process is errorprone, and goes against the spirit of LATEX. Instead, follow my advice on page 400, and let LATEX do the work for you. In the preamble, type the proclamations \newtheorem{theorem}{Theorem} \newtheorem{corollary}{Corollary}[theorem] We are almost there. Theorem 1 now is followed by Corollary 1.1, Corollary 1.2 and Theorem 3 by Corollary 3.1. If we redefine \thecorollary, \renewcommand{\thecorollary}{\arabic{corollary}} then Theorem 1 is followed by Corollary 1 and Corollary 2, and Theorem 3 is also
15.5
Numbering and measuring
403
followed by Corollary 1. If you need to perform more complicated arithmetic with counters, use Kresten K. Thorup and Frank Jensen’s calc package (see Section 12.3.1). This package is discussed in Section A.3.1 of The LATEX Companion, 2nd edition [46]. Two special counters The secnumdepth and tocdepth counters control which sectional units are numbered and which are listed in the table of contents, respectively. For example, \setcounter{secnumdepth}{2} sets secnumdepth to 2. As a result, chapters—if they are present in the document class—sections, and subsections are numbered, but subsubsections are not. This command must be placed in the preamble of the document. tocdepth is similar.
15.5.2 Length commands While a counter contains integers, a length command contains a real number and a dimensional unit. LATEX recognizes many different dimensional units. We list five absolute units: cm centimeter in inch pc pica (1 pc = 12 pt) pt point (1 in = 72.27 pt) mm millimeter and two relative units: em, approximately the width of the letter M in the current font ex, approximately the height of the letter x in the current font LATEX defines many length commands. For instance, Section 4.1 of The LATEX Companion, 2nd edition [46] lists 17 length commands for page layout alone. You can find some of them in Figure 10.4. A list environment sets about a dozen additional length commands (see Figure 15.2). Length commands are defined for almost every aspect of LATEX’s work, including displayed math environments—a complete list would probably contain a few hundred. Many are listed in Leslie Lamport’s LATEX: A Document Preparation System, 2nd edition [43] and in The LATEX Companion, 2nd edition [46]. Many more are hidden in packages such as amsmath.
404
Chapter 15
Customizing LATEX
The most common length commands are: \parindent, the amount of indentation at the beginning of a paragraph \parskip, the extra vertical space inserted between paragraphs \textwidth, the width of the text on a page A more esoteric example is \marginparpush, the minimum vertical space between two marginal notes. Luckily, you do not have to be familiar with many length commands because LATEX and the document class set them for you. Defining new length commands You can define your own length commands. For example, \newlength{\mylength} makes \mylength a new length command with a value of 0 points. Note that while you have to type \newcounter{counter} to get a new counter, typing \newlength{mylength} results in an error message such as ! Missing control sequence inserted. \inaccessible l.3 \newlength{mylength} Setting length The \setlength command sets or resets the value of a length command. So \setlength{\textwidth}{3in} creates a very narrow page. The first argument of \setlength must be a length command, not simply the command name, that is \setlength{textwidth}{3in} % Bad is incorrect. The second argument of \setlength must be a real number with a dimensional unit, for instance, 3in, and not simply a real number. In other words, \setlength{\textwidth}{3} % Bad is also incorrect.
15.5
405
Numbering and measuring
Tip
A common mistake is to type a command such as \setlength{\marginpar}{0} Instead, type \setlength{\marginpar}{0pt} Always be sure to include a dimensional unit. The \addtolength command adds a quantity to the value of a length command. For instance, \addtolength{\textwidth}{10pt} narrows the page width by 10 points. If you define \newlength{\shorterlength} \setlength{\shorterlength}{\mylength} \addtolength{\shorterlength}{.5in} then \parbox{\shorterlength}{...} always typesets its second argument in a box 1/2 inch narrower than the parboxes set to be of width \mylength. When LATEX typesets some text or math, it creates a box. Three measurements are used to describe the size of the box: the width the height, from the baseline to the top the depth, from the baseline to the bottom as illustrated in Figure 15.1. For instance, the box typesetting “aa” has a width of 10.00003 pt, a height of 4.30554 pt, and a depth of 0 pt. The box typesetting “ag” has
Figure 15.1: The measurements of a box.
406
Chapter 15
Customizing LATEX
the same width and height, but a depth of 1.94444 pt. The box “Ag” (see Figure 15.1) has a width of 12.50003 pt, a height of 6.83331 pt, and a depth of 1.94444 pt. The commands \settowidth \settoheight \settodepth each take two arguments. The first argument is a length command, the second is text (or math) to be measured by LATEX. The corresponding measurement of the box in which the second argument is typeset is assigned to the length command in the first argument. For example, if \mylength is a length command, then \settowidth{\mylength}{Ag} assigns 12.50003 pt to \mylength. It should be clear from this example how the \phantom and \hphantom commands (see Section 5.8.1) are related to this command. To perform more complicated arithmetic with length commands, use the calc package. Rubber lengths In addition to rigid lengths, such as 3in, LATEX can also set rubber lengths, that is, lengths that are allowed to stretch and shrink. Here is an example: \setlength{\stretchspace}{3in plus 10pt minus 8pt} Assuming that \stretchspace is a length command, this command assigns it a value of 3 inches that can stretch by 10 points or shrink by 8 points, if necessary. So a box of width \stretchspace is 3 inches wide, plus up to 10 points, or minus up to 8 points. Stretchable vertical spaces are often used before and after displayed text environments. LATEX adjusts these spaces to make the page look balanced. An example can be found in Section 15.1.7. \medskipamount is defined as 6.0pt plus 2.0pt minus 2.0pt See Section 15.6.3 for more examples. The \fill command is a special rubber length that can stretch any amount. The stretching is done evenly if there is more than one \fill present. See the second example of brandnew environments in Section 15.2.5.
15.6
Custom lists Although there are three readymade list environments provided by LATEX (see Section 6.2), it is often necessary to create one of your own using LATEX’s list environment. In fact, LATEX itself uses the list environment to define many of its standard environments, including:
15.6
Custom lists
407
The three list environments (Section 6.2) The quote, quotation, and verse environments (Section 6.8) Proclamations (Section 6.4) The style environments center, flushleft, and flushright (Section 6.3) The thebibliography environment (Section 10.5.1) The theindex environment (Section 10.5.2)
15.6.1 Length commands for the list environment The general layout of a list is shown in Figure 15.2. It uses six horizontal measurements and three vertical measurements. I now list these length commands. Vertical length commands \topsep is most of the vertical space between the first item and the preceding text, and also between the last item and the following text. This space also includes \parskip, the extra vertical space inserted between paragraphs, and optionally, \partopsep, provided that the list environment starts a new paragraph. \parsep is the space between paragraphs of the same item. \itemsep is the space between items. Like \topsep, the actual gap is the sum of \itemsep and \parsep. All of these vertical length commands are rubber lengths (see Section 15.5.2). Horizontal length commands By default, the margins of a list environment are the same as the margins of the surrounding text. If the list is nested within a list, the margins are wider and so the text is narrower. The \leftmargin and \rightmargin length commands specify the distance between the edge of the item box and the left and right margins of the page. The label is the text provided by the optional argument of an \item command or provided as a default in the definition of the list environment. It is typeset in a box of width \labelwidth, which is indented \itemindent units from the left margin, and separated by a space of \labelsep units from the text box. If the label is too wide to fit in the box, it is typeset at its full natural width, and the first line in the text box is indented. The second and subsequent paragraphs of an item are typeset with their first lines indented by \listparindent units.
408
Chapter 15
Customizing LATEX
Surrounding text
\itemindent
\topsep+\parskip
\labelsep
label Item 1 paragraph 1
\labelwidth
\rightmargin
\leftmargin \listparindent
\parsep
Item 1 paragraph 2
\itemsep+\parsep
label Item 2 paragraph 1
\topsep+\parskip
Surrounding text
Figure 15.2: The layout of a custom list.
15.6
Custom lists
409
15.6.2 The list environment Custom lists are created with the list environment, which is invoked as follows: \begin{list}{default_label }{declarations } \item item1 \item item2 ... \end{list} The arguments are default label, the label for any items that do not specify their own, similar to the optional argument of the \item command declarations, the vertical and horizontal length commands and any other required parameters for the list Here is a very simple example: Here are the most important LATEX rules about spaces in text, sentences, and paragraphs: ♦ Rule 1: Two or more spaces in text are the same as one. ♦ Rule 2: A blank line (that is, two endofline characters separated only by blanks and tabs) indicates the end of a paragraph. Rules 1 and 2 make typing and copying very convenient. I have used the ♦ math symbol ($\diamondsuit$) as a default label, and I set the item box 0.5 inch from either margin. So this example is typed as follows: \noindent Here are the most important \LaTeX\ rules about spaces in text, sentences, and paragraphs: \begin{list}{$\diamondsuit$}{\setlength{\leftmargin}% {.5in}\setlength{\rightmargin}{.5in}} \item \textbf{Rule 1:} Two or more spaces in text are the same as one. \item \textbf{Rule 2:} A blank line (that is, two endofline characters separated only by blanks and tabs) indicates the end of a paragraph. \end{list} Rules 1 and~2 make typing and copying very convenient. Here is a second variant:
410
Chapter 15
Customizing LATEX
Here are the most important LATEX rules about spaces in text, sentences, and paragraphs: Rule 1: Two or more consecutive spaces in text are the same as one. Rule 2: A blank line (that is, two endofline characters separated only by blanks and tabs) indicates the end of a paragraph. Rules 1 and 2 make typing and copying very convenient. In this example, I dropped the optional default label and typed Rule 1: and Rule 2: as (optional) arguments of the \item commands: \noindent Here are the most important \LaTeX\ rules about spaces in text, sentences, and paragraphs: \begin{list}{}{\setlength{\leftmargin}{.5in}% \setlength{\rightmargin}{.5in}} \item[\textbf{Rule 1:}] Two or more consecutive spaces in text are the same as one. \item[\textbf{Rule 2:}] A blank line (that is, two endofline characters separated only by blanks and tabs) indicates the end of a paragraph. \end{list} Rules 1 and~2 make typing and copying very convenient. For further simple examples, you can look at various document class files to see how standard environments such as verse, quote, and so on, are defined. Using counters It is not very LATEXlike to provide the numbers for the rules in the examples above. It would be more logical for LATEX to do the numbering. The following is a more LATEXlike coding of the second example: \noindent Here are the most important \LaTeX\ rules about spaces in text, sentences, and paragraphs: \newcounter{spacerule} \begin{list}{\textbf{Rule \arabic{spacerule}:}} {\setlength{\leftmargin}{.5in} \setlength{\rightmargin}{.5in} \usecounter{spacerule}} \item Two or more consecutive spaces in text are the same as one.\label{Li:Twoor} \item A blank line (that is, two endofline
15.6
411
Custom lists characters separated only by blanks and tabs) indicates the end of a paragraph.\label{Li:blankline} \end{list} Rules \ref{Li:Twoor} and~\ref{Li:blankline} make typing and copying very convenient. Note that 1. I declared the counter before the list environment with the line \newcounter{spacerule} 2. I defined the default label as
\textbf{Rule \arabic{spacerule}:} item In the declarations, I specified that the list should use the spacerule counter with the command \usecounter{spacerule}
15.6.3 Two complete examples In the previous examples, I set the values of \leftmargin and \rightmargin. The other length commands were not redefined, so their values remained the values set by the document class. In the following examples, I set the values of many more length commands. Example 1 To get the following list, Here are the most important LATEX rules about spaces in text, sentences, and paragraphs: Rule 1:
Two or more consecutive spaces in text are the same as one.
Rule 2:
A blank line—that is, two endofline characters separated only by blanks and tabs— indicates the end of a paragraph.
Rules 1 and 2 make typing and copying very convenient. we type \noindent Here are the most important \LaTeX\ rules about spaces in text, sentences, and paragraphs: \newcounter{spacerule} \begin{list}{\upshape\bfseries Rule \arabic{spacerule}:}
412
Chapter 15
Customizing LATEX
{\setlength{\leftmargin}{1.5in} \setlength{\rightmargin}{0.6in} \setlength{\labelwidth}{1.0in} \setlength{\labelsep}{0.2in} \setlength{\parsep}{0.5ex plus 0.2ex minus 0.1ex} \setlength{\itemsep}{0ex plus 0.2ex minus 0ex} \usecounter{spacerule} \itshape} \item Two or more consecutive spaces in text are the same as one.\label{Li:Twoor} \item A blank linethat is, two endofline characters separated only by blanks and tabsindicates the end of a paragraph.\label{Li:blankline} \end{list} Rules \ref{Li:Twoor} and~\ref{Li:blankline} make typing and copying very convenient. Note that 1. I declared the counter as in the previous example. 2. The last item in declarations is \itshape, which typesets the entire list in italics. 3. The default label is defined as \upshape\bfseries Rule \arabic{spacerule} My first attempt was to define it as \bfseries Rule \arabic{spacerule} which typesets Rule in bold italics (because in Step 2 we set the whole list in italics). To force the label to be typeset upright, I start the default label with the \upshape command. 4. The left margin is set to 1.5 inches and the right margin to 0.6 inches: \setlength{\leftmargin}{1.5in} \setlength{\rightmargin}{0.6in} 5. Next I set the width of the label to 1 inch, and the space between the label and the item to 0.2 inches:
15.6
413
Custom lists \setlength{\labelwidth}{1.0in} \setlength{\labelsep}{0.2in}
6. Finally, I set the paragraph separation to 0.5 ex, allowing stretching by 0.2 ex and shrinking by 0.1 ex, and the item separation to 0 ex, allowing stretching by 0.2 ex and no shrinking, by \setlength{\parsep}{0.5ex plus 0.2ex minus 0.1ex} \setlength{\itemsep}{0ex plus 0.2ex minus 0ex} The actual amount of item separation is calculated by adding the values specified for \parsep and \itemsep. A complicated list such as this should be defined as a new environment. For example, you could define a myrules environment: \newenvironment{myrules} {\begin{list} {\upshape \bfseries Rule \arabic{spacerule}:} {\setlength{\leftmargin}{1.5in} \setlength{\rightmargin}{0.6in} \setlength{\labelwidth}{1.0in} \setlength{\labelsep}{0.2in} \setlength{\parsep}{0.5ex plus 0.2ex minus 0.1ex} \setlength{\itemsep}{0ex plus 0.2ex minus 0ex} \usecounter{spacerule} \itshape} } {\end{list}} and then use it anywhere, as in \begin{myrules} \item Two or more consecutive spaces in text are the same as one.\label{Li:Twoor} \item A blank linethat is, two endofline characters separated only by blanks and tabsindicates the end of a paragraph. \label{Li:blankline} \end{myrules} Rules \ref{Li:Twoor} and~\ref{Li:blankline} make typing and copying very convenient. which typesets as the first example shown on page 411.
414
Chapter 15
Customizing LATEX
Example 2 In Section 5.7.2, we discussed the formatting of the following type of glossary: sentence is a group of words terminated by a period, exclamation point, or question mark. paragraph is a group of sentences terminated by a blank line or by the \par command. Now we can create the glossary as a custom list: \begin{list}{} {\setlength{\leftmargin}{30pt} \setlength{\rightmargin}{0pt} \setlength{\itemindent}{14pt} \setlength{\labelwidth}{40pt} \setlength{\labelsep}{5pt} \setlength{\parsep}{0.5ex plus 0.2ex minus 0.1ex} \setlength{\itemsep}{0ex plus 0.2ex minus 0ex}} \item[\textbf{sentence}\hfill] is a group of words terminated by a period, exclamation point, or question mark. \item[\textbf{paragraph}\hfill] is a group of sentences terminated by a blank line or by the \com{par} command. \end{list} There is nothing new in this example except the \hfill commands in the optional arguments to left adjust the labels. With the long words in the example this adjustment is not necessary, but it would be needed for shorter words. See Section 3.3 of The LATEX Companion, 2nd edition [46] on how to customize the three standard list environments and also for more complicated custom lists.
15.6.4
The trivlist environment LATEX also provides a trivlist environment, meant more for programmers than users. The environment is invoked in the form \begin{trivlist} body \end{trivlist} It is similar to the list environment except that there are no arguments, and all the length commands are trivially set, most to 0 points, except for \listparindent and \parsep, which are set to equal \parindent and \parskip, respectively. For instance, LATEX defines the center environment as follows:
15.7
The dangers of customization
415
\begin{trivlist} \centering \item[] \end{trivlist}
15.7
The dangers of customization We can customize LATEX in so many ways. We can add packages to expand its power and define new commands that better suit our work habits. These enhance LATEX and make it easier to work with. But they also introduce difficulties. Let us start with the obvious. Whoever introduced the command \textcompwordmark knew that—even if we use command completion—we are not going to type if\textcompwordmark f to avoid having a ligature (see Section 5.4.6). It is a lot of typing, and the source file becomes hard to read. This cries out for a userdefined command, say, \Iff, which is short and readable (see Section 15.1.1). When introducing userdefined commands, watch out for the following traps.
Trap 1
Redefining a command that is a necessary part of LATEX. This is easy to avoid. As discussed in Section 15.1.7, you can easily find out whether a command is already in use. If it is, do not redefine it unless you really know what you are doing.
Trap 2
Defining too many commands. This creates two problems. Your editor has a hard time making changes in your source file. And a few years later, when you want to reuse the material, you have a difficult time understanding all those clever commands.
Trap 3
Your contribution appears in a volume with many other authors and your userdefined commands create conflicts. As your article appears in a publication, some parts of it are used for the whole volume. The title and maybe even the section titles are used in the table of contents. The abstracts may be collected for the whole volume or there may be a joint bibliography.
416
Chapter 15
Customizing LATEX
With the advent of the internet, there are now collections of thousands of math articles; PlanetMath.org is one example. Write your articles so that even the editors of PlanetMath.org can use it.
Rule 1
Do not use your own commands in the title of the article, in the abstract, in section titles, in the bibliography, or in captions of figures and tables.
Trap 4
You submit the article to a journal that does not permit a separate custom command file. For such journals, just copy the needed userdefined commands into the preamble of your article. Go through the list and delete those userdefined commands that are not used in this article. This helps the editor to look up your commands from a shorter list.
Rule 2
Introduce judiciously userdefined commands with very short names. Introducing oneletter commands—for instance, using \C for the complex field— is dangerous because many oneletter commands are reserved by LATEX. Twoletter userdefined commands are not quite this bad. Of the 2,500 or so possibilities only a few dozen are used by LATEX. The danger here is, of course, conflict with other authors and confusion for the editor. My command file has about 15 twoletter commands. For instance \jj, part of the \jj, \JJ, \JJm family. Also \Id, because Id is the standard notation for ideal lattices. Some editors may think that this is 15 too many.
Rule 3
Do not use \def to define your commands, with the exception of a very few delimited commands. Using \def means giving up LATEX’s built in defense. In the editorial office of my journal, about half the submitted articles that we cannot typeset violate this rule.
Rule 4
Do not redefine length commands, especially, if you do not know what other length commands are computed based on the ones you change.
15.7
The dangers of customization
417
The page layout diagram, Figure 10.4, should provide examples. Even simpler, Do not redefine length commands. Let the document class define them for your article.
Rule 5
Make sure that the packages you use are compatible. For instance, the popular psfrag and epsfig packages cause problems if used with the AMS packages. Be cautious when you use packages that redefine a lot of LATEX commands, such as hyperref (see Section 13.2). You can read more about the plight of authors in the hands of incompetent editors in my article [31] and the difficult job of editors with articles violating the above rules in Enrico Gregorio [16].
CHAPTER
16 BIBTEX
The B IBTEX application, written by Oren Patashnik, assists LATEX users in compiling bibliographies, especially long ones. Short bibliographies can easily be placed in the document directly (see Section 10.5.1). It takes a little effort to learn B IBTEX. But in the long run, the advantages of building bibliographic databases that can be reused and shared outweigh the disadvantage of a somewhat steep learning curve. The bibliographic database files, the bib files, contain the bibliographic entries. We discuss the format of these entries in Section 16.1, and then describe how to use B IBTEX to create bibliographies in Section 16.2. B IBTEX uses a style, called a bibliographic style, or bst file, to format entries. On the next two pages we show the bibliography of the sampartb.tex sample article typeset with six different style files. To simplify our discussion, in the rest of this chapter I discuss only one style, the AMS plain style, amsplain.bst, version 2.0. All of the examples shown are in this style, and several of the comments I make are true only for the AMS plain style. If you choose to use a different style, you should check its documentation for special rules.
422
Chapter 16
BIBTEX
[1] SooKey Foo. Lattice Constructions. PhD thesis, University of Winnebago, Winnebago, MN, December 1990. [2] George A. Menuhin. Universal Algebra. D. van Nostrand, Princeton, 1968. [3] Ernest T. Moynahan. Ideals and congruence relations in lattices. II. Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl., 7:417–434, 1957. [4] Ernest T. Moynahan. On a problem of M. Stone. Acta Math. Acad. Sci. Hungar., 8:455–460, 1957. [5] Ferenc R. Richardson. General Lattice Theory. Mir, Moscow, expanded and revised edition, 1982.
plain.bst
[Foo90]
SooKey Foo. Lattice Constructions. PhD thesis, University of Winnebago, Winnebago, MN, December 1990. [Men68] George A. Menuhin. Universal Algebra. D. van Nostrand, Princeton, 1968. [Moy57a] Ernest T. Moynahan. Ideals and congruence relations in lattices. II. Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl., 7:417–434, 1957. [Moy57b] Ernest T. Moynahan. On a problem of M. Stone. Acta Math. Acad. Sci. Hungar., 8:455–460, 1957. [Ric82] Ferenc R. Richardson. General Lattice Theory. Mir, Moscow, expanded and revised edition, 1982.
alpha.bst
1. SooKey Foo, Lattice constructions, Ph.D. thesis, University of Winnebago, Winnebago, MN, December 1990. 2. George A. Menuhin, Universal algebra, D. van Nostrand, Princeton, 1968. 3. Ernest T. Moynahan, Ideals and congruence relations in lattices. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl. 7 (1957), 417–434 (Hungarian). 4. , On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. 5. Ferenc R. Richardson, General lattice theory, expanded and revised ed., Mir, Moscow, 1982 (Russian).
amsplain.bst [Foo90]
SooKey Foo, Lattice constructions, Ph.D. thesis, University of Winnebago, Winnebago, MN, December 1990. [Men68] George A. Menuhin, Universal algebra, D. van Nostrand, Princeton, 1968. [Moy57a] Ernest T. Moynahan, Ideals and congruence relations in lattices. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl. 7 (1957), 417–434 (Hungarian). [Moy57b] Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. [Ric82] Ferenc R. Richardson, General lattice theory, expanded and revised ed., Mir, Moscow, 1982 (Russian).
amsalpha.bst
16.1
423
The database
[1] S.K. Foo, Lattice Constructions, PhD thesis, University of Winnebago, Winnebago, MN, Dec. 1990. [2] G. A. Menuhin, Universal Algebra, D. van Nostrand, Princeton, 1968. [3] E. T. Moynahan, Ideals and congruence relations in lattices. II, Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl., 7 (1957), pp. 417–434. [4] , On a problem of M. Stone, Acta Math. Acad. Sci. Hungar., 8 (1957), pp. 455–460. [5] F. R. Richardson, General Lattice Theory, Mir, Moscow, expanded and revised ed., 1982.
siam.bst [1] F. R. Richardson, General Lattice Theory. Moscow: Mir, expanded and revised ed., 1982. [2] G. A. Menuhin, Universal Algebra. Princeton: D. van Nostrand, 1968. [3] E. T. Moynahan, “On a problem of M. Stone,” Acta Math. Acad. Sci. Hungar., vol. 8, pp. 455– 460, 1957. [4] S.K. Foo, Lattice Constructions. PhD thesis, University of Winnebago, Winnebago, MN, Dec. 1990. [5] E. T. Moynahan, “Ideals and congruence relations in lattices. II,” Magyar Tud. Akad. Mat. Fiz. Oszt. K¨ ozl., vol. 7, pp. 417–434, 1957.
ieeetr.bst
16.1
The database A B IBTEX database is a text file containing bibliographic entries. To use B IBTEX, you first have to learn how to assemble a database. This section explains how to do that. There may be special tools available for your computer system that assist you in building and maintaining your bibliographic data. Such tools make compiling the data easier and may minimize formatting errors. You can find all the examples in this section in the template.bib file in the samples folder (see page 4).
16.1.1 Entry types A bibliographic entry is given in pieces called fields. The style (see Section 16.2.2) specifies how these fields are typeset. Here are two typical entries: @BOOK{gM68, author = "George A. Menuhin", title = "Universal Algebra", publisher = "D.~Van Nostrand", address = "Princeton", year = 1968, }
424
Chapter 16
BIBTEX
@ARTICLE{eM57, author = "Ernest T. Moynahan", title = "On a Problem of {M. Stone}", journal = "Acta Math. Acad. Sci. Hungar.", pages = "455460", volume = 8, year = 1957, } The start of an entry is indicated with an at sign (@) followed by the entry type. In the first example, the entry type is BOOK, while in the second, it is ARTICLE. The entry type is followed by a left brace ({). The matching right brace (}) indicates the end of the entry. B IBTEX also allows you to use parentheses as delimiters for an entry. In this book, however, we use braces to enclose an entry. The string @BOOK{ is followed by a label, gM68, which designates the name of the entry. Refer to this entry in your document with \cite{gM68}. The label is followed by a comma and a series of fields. In this example, there are five fields, author, title, publisher, address, and year. Each field starts with the field name, followed by = and the value of the field enclosed in double quotes ("). Be sure to use " and not LATEX double quotes (‘‘ or ’’). Alternatively, B IBTEX also allows you to use braces to enclose the field value. In this book, we use double quotes to enclose a field. Numeric field values, that is, fields consisting entirely of digits, do not need to be enclosed in double quotes or braces, for instance, year in the examples above, volume in the second example, and number in some of the examples that follow. Page ranges, such as 455460, are not numeric field values since they contain , so they must be enclosed in double quotes or braces. There must be a comma before each field. The comma before the first field is placed after the label. There are many standard entry types, including ARTICLE an article in a journal or magazine BOOK a book with an author (or editor) and a publisher BOOKLET a printed work without a publisher INBOOK a part of a book, such as a chapter or a page range that, in general, is not titled or authored separately INCOLLECTION a part of a book with its own title and perhaps author INPROCEEDINGS an article in a conference proceedings with its own title and author MANUAL technical documentation
16.1
425
The database MASTERSTHESIS a master’s thesis MISC an entry that does not fit in any other category PHDTHESIS a Ph.D. thesis PROCEEDINGS the proceedings of a conference TECHREPORT a report published by a school or institution UNPUBLISHED an unpublished paper Each entry includes a number of fields from the following list: address author booktitle chapter crossref edition editor howpublished
institution journal key language month note number organization
pages publisher school series title type volume year
The style you choose determines which of the fields within an entry are actually used. All the others are ignored. You may also add fields for your own use. For example, you may want to add a mycomments field for personal comments. Such fields are ignored unless you have a bibliography style that uses them. Commonly used examples of new field names include URL, abstract, ISBN, keywords, mrnumber, and so on. The language field is used by the AMS styles but not by any of the other styles mentioned in this chapter.
Tip 1. B IBTEX does not care whether you use uppercase or lowercase letters (or mixed) for the names of entry types and fields. In this book, the entry types are shown in uppercase and field names in lowercase. 2. Placing a comma after the last field is optional. I recommend that you put it there so that when you append a new field to the entry, the required comma separating the fields is present.
For each entry type there are both required and optional fields. Later in this section, I give two examples of each entry type. The first example of an entry type uses a small set of fields, while the second example is a maximal one, showing a large number of optional fields.
426
16.1.2
Chapter 16
BIBTEX
Typing fields Make sure you type the field names correctly. If you misspell one, B IBTEX ignores the field. B IBTEX also warns you if a required field is missing. The author and editor fields require a name.
Rule
Names 1. Most names can be typed as usual, "Ernest T. Moynahan" or "Moynahan, Ernest T.", with one comma separating the family name from the given names. 2. Type two or more names separated by and. For instance, author= "George Blue and Ernest Brown and SooKey Foo", 3. The family name of Miguel Lopez Fernandez is Lopez Fernandez, so type it as "Lopez Fernandez, Miguel". This informs B IBTEX that Lopez is not a middle name. 4. Type Orrin Frink, Jr. as "Frink, Jr., Orrin". Rules 3 and 4 are seldom needed. In a bibliography of about 1,500 items, I found fewer than 10 names that could not be typed as usual. Note that you can type John von Neumann as "John von Neumann" or "von Neumann, John". Because B IBTEX knows about von, it handles the name properly. There are a few rules concerning the title field.
Rule
Title 1. You should not put a period at the end of a title. The style supplies the appropriate punctuation. 2. Many styles, including the AMS styles, convert titles, except for the first letter of the title, to lowercase for all entry types. If you want a letter to appear in uppercase, put it—or the entire word—in braces. The same rule applies to the edition field. Some other styles only do this conversion for the titles of nonbooklike entries. 3. To maximize the portability of your database, you should type titles with each important word capitalized: title = "On a Problem of {M. Stone}", The style used in this book, amsplain.bst, converts Problem to problem, so it
16.1
The database
427
makes no difference, but some styles do not. To be on the safe side, you should capitalize all words that may have to be capitalized. For the record, here are the complete rules for titles: Capitalize (1) the first word; (2) the first word in a subtitle (B IBTEX assumes that a subtitle follows a colon, so it capitalizes the first word after a colon—a colon not introducing a subtitle should be typed in braces); (3) all other words except articles, unstressed conjunctions, and unstressed prepositions. Words that should never be converted to lowercase, for example proper names such as Hilbert, should be enclosed in braces to prevent them from being converted to lowercase. In the example above, two letters in the title should not be converted to lowercase, so we enclosed M. Stone in braces. We could also have typed {M. S}tone or {M.} {S}tone. B IBTEX and the style automatically handle a number of things for you that you would have to handle yourself when typing text. 1. You do not have to mark periods in abbreviations, as .\ in the names of journals (see Section 5.2.2). So journal = "Acta Math. Acad. Sci. Hungar.", typesets correctly. 2. You can type a single hyphen for a page range instead of the usual  in the pages field (see Section 5.4.2). So pages = "455460", typesets correctly with an en dash. 3. You do not have to type nonbreakable spaces with ~ in the author or editor fields (see Section 5.4.3): author = "George A. Menuhin", is correct. Normally you would type George~A. Menuhin. Finally, we state a rule about accented characters.
Rule
Accents Put accented characters in braces: {\"{a}}. This rule means that author = "Paul Erd\H{o}s",
428
Chapter 16
BIBTEX
is not recommended. Instead, type author = "Paul Erd{\H{o}}s", This rule is, again, about portability. Some styles, e.g., alpha and amsalpha, create a citation for an article from the first three letters of the name and the last two digits of the year. author = "Kurt G{\"{o}}del", year = 1931, creates the citation: [G¨ od31]. The accent is used only if the accents rule has been followed. The downside of this rule is that the braces suppress kerning.
16.1.3
Articles Entry type Required fields Optional fields
ARTICLE author, title, journal, year, pages volume, number, language, note
Examples: 1. Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460. 2. Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), no. 5, 455–460 (English), Russian translation available. typed as @ARTICLE{eM57, author = "Ernest T. Moynahan", title = "On a Problem of {M. Stone}", journal = "Acta Math. Acad. Sci. Hungar.", pages = "455460", volume = 8, year = 1957, } @ARTICLE{eM57a, author = "Ernest T. Moynahan", title = "On a Problem of {M. Stone}", journal = "Acta Math. Acad. Sci. Hungar.", pages = "455460",
16.1
429
The database volume = 8, number = 5, year = 1957, note = "Russian translation available", language = "English", }
16.1.4 Books Entry type Required fields Optional fields
BOOK author (or editor), title, publisher, year edition, series, volume, number, address, month, language, note
Examples: 1. George A. Menuhin, Universal algebra, D. Van Nostrand, Princeton, 1968. 2. George A. Menuhin, Universal algebra, second ed., University Series in Higher Mathematics, vol. 58, D. Van Nostrand, Princeton, March 1968 (English), no Russian translation. typed as @BOOK{gM68, author = "George A. Menuhin", title = "Universal Algebra", publisher = "D.~Van Nostrand", address = "Princeton", year = 1968, } @BOOK{gM68a, author = "George A. Menuhin", title = "Universal Algebra", publisher = "D.~Van Nostrand", address = "Princeton", year = 1968, month = mar, series = "University Series in Higher Mathematics", volume = 58, edition = "Second", note = "no Russian translation", language = "English", }
430
Chapter 16
BIBTEX
Abbreviations, such as mar, are discussed in Section 16.1.9. A second variant of book has an editor instead of an author: 15. Robert S. Prescott (ed.), Universal algebra, D. Van Nostrand, Princeton, 1968. typed as @BOOK{rP68, editor = "Robert S. Prescott", title = "Universal Algebra", publisher = "D.~Van Nostrand", address = "Princeton", year = 1968, }
16.1.5
Conference proceedings and collections Entry type Required fields Optional fields
INPROCEEDINGS author, title, booktitle, year address, editor, series, volume, number, organization, publisher, month, note, pages, language
Examples: 7. Peter A. Konig, Composition of functions. Proceedings of the Conference on Universal Algebra, 1970. 8. Peter A. Konig, Composition of functions. Proceedings of the Conference on Universal Algebra (Kingston, ON) (G. H. Birnbaum, ed.), vol. 7, Canadian Mathematical Society, Queen’s Univ., December 1970, available from the Montreal office, pp. 1–106 (English). typed as @INPROCEEDINGS{pK69, author = "Peter A. Konig", title = "Composition of Functions". booktitle = "Proceedings of the Conference on Universal Algebra", year = 1970, }
16.1
431
The database @INPROCEEDINGS{pK69a, author = "Peter A. Konig", title = "Composition of Functions". booktitle = "Proceedings of the Conference on Universal Algebra", address = "Kingston, ON", publisher = "Queen’s Univ.", organization = "Canadian Mathematical Society", editor = "G. H. Birnbaum", pages = "1106", volume = 7, year = 1970, month = dec, note = "available from the Montreal office", language = "English", }
The address field provides the location of the meeting. The address of the publisher should be in the publisher field and the address of the organization in the organization field. Entry type Required fields Optional fields
INCOLLECTION author, title, booktitle, publisher, year editor, series, volume, number, address, edition, month, note, pages, language
Examples: 1. Henry H. Albert, Free torsoids, Current Trends in Lattices, D. Van Nostrand, 1970. 2. Henry H. Albert, Free torsoids, Current Trends in Lattices (George Burns, ed.), vol. 2, D. Van Nostrand, Princeton, January 1970, new edition is due next year, pp. 173–215 (German). is typed as @INCOLLECTION{hA70, author = "Henry H. Albert", title = "Free Torsoids", booktitle = "Current Trends in Lattices". publisher = "D.~Van Nostrand", year = 1970, }
432
Chapter 16
BIBTEX
@INCOLLECTION{hA70a, author = "Henry H. Albert", editor = "George Burns", title = "Free Torsoids", booktitle = "Current Trends in Lattices". publisher = "D.~Van Nostrand", address = "Princeton", pages = "173215", volume = 2, year = 1970, month = jan, note = "new edition is due next year", language = "German", } The address field contains the address of the publisher. Crossreferencing If your database has several articles from the same conference proceedings and collections, you may prefer to make an entry for the entire volume, and crossreference individual articles to that entry. For instance, @PROCEEDINGS{UA69, title = "Proceedings of the Conference on, Universal Algebra", booktitle = "Proceedings of the Conference on Universal Algebra", address = "Kingston, ON", publisher = "Canadian Mathematical Society", editor = "G. H. Birnbaum", volume = 7, year = 1970, } may be the entry for the proceedings volume as a whole, and @INPROCEEDINGS{pK69a, author = "Peter A. Konig", title = "Composition of Functions", booktitle = "Proceedings of the Conference on Universal Algebra", pages = "1106", crossref = "UA69", }
16.1
433
The database
is the crossreferencing entry for a specific article. These two entries produce the following: 1. G. H. Birnbaum (ed.), Proceedings of the conference on universal algebra, vol. 7, Kingston, ON, Canadian Mathematical Society, 1970. 2. Peter A. Konig, Composition of functions, in Birnbaum [1], pp. 1–106.
Rule
Crossreferences 1. All the required fields of the crossreferencing entry must appear in either that entry or in the crossreferenced entry. 2. The crossreferenced entry should have both a title and a booktitle field. 3. The crossreferenced entry must appear in the bib file later than any entry that crossreferences it.
16.1.6 Theses Entry type Required fields Optional fields Examples:
MASTERSTHESIS or PHDTHESIS author, title, school, year type, address, month, note, pages
1. SooKey Foo, Lattice constructions, Ph.D. thesis, University of Winnebago, 1990. 2. SooKey Foo, Lattice constructions, Ph.D. dissertation, University of Winnebago, Winnebago, MN, December 1990, final revision not yet available, pp. 1–126. is typed as @PHDTHESIS{sF90, author = "SooKey Foo", title = "Lattice Constructions", school = "University of Winnebago", year = 1990, }
434
Chapter 16
BIBTEX
@PHDTHESIS{sF90a, author = "SooKey Foo", title = "Lattice Constructions", school = "University of Winnebago", address = "Winnebago, MN", year = 1990, month = dec, note = "final revision not yet available", type = "Ph.D. dissertation", pages = "1126", } If the type field is present, its content takes the place of the phrase Ph.D. thesis (or Master’s thesis).
16.1.7
Technical reports Entry type Required fields Optional fields Examples:
TECHREPORT author, title, institution, year type, number, address, month, note
1. Grant H. Foster, Computational complexity in lattice theory, tech. report, Carnegie Mellon University, 1986. 2. Grant H. Foster, Computational complexity in lattice theory, Research Note 128A, Carnegie Mellon University, Pittsburgh, PA, December 1986, in preparation. is typed as @TECHREPORT{gF86, author = "Grant H. Foster", title = "Computational Complexity in Lattice Theory", institution = "Carnegie Mellon University", year = 1986, } @TECHREPORT{gF86a, author = "Grant H. Foster", title = "Computational Complexity in Lattice Theory", institution = "Carnegie Mellon University", year = 1986,
16.1
435
The database month = dec, type = "Research Note", address = "Pittsburgh, PA", number = "128A", note = "in preparation", }
16.1.8 Manuscripts and other entry types Entry type Required fields Optional fields
UNPUBLISHED author, title, note month, year
Examples: 1. William A. Landau, Representations of complete lattices, manuscript, 55 pages. 2. William A. Landau, Representations of complete lattices, manuscript, 55 pages, December 1975. is typed as @UNPUBLISHED{wL75, author = "William A. Landau", title = "Representations of Complete Lattices", note = "manuscript, 55~pages", } @UNPUBLISHED{wL75a, author = "William A. Landau", title = "Representations of Complete Lattices", year = 1975, month = dec, note = "manuscript, 55~pages", } Other standard entry types include Entry type BOOKLET Required field title Optional fields author, howpublished, address, month, year, note
436
Chapter 16
BIBTEX
Entry type INBOOK Required fields author or editor, title, chapter or pages, publisher, year Optional fields series, volume, number, type, address, edition, month, pages, language, note Entry type MANUAL Required field title Optional fields author, organization, address, edition, month, year, note Entry type MISC Required field at least one of the optional fields must be present Optional fields author, title, howpublished, month, year, note, pages Entry type PROCEEDINGS Required fields title, year Optional fields editor, series, volume, number, address, organization, publisher, month, note
16.1.9
Abbreviations You may have noticed the field month = dec in some of the examples. This field uses an abbreviation. Most B IBTEX styles, including the AMS styles, include abbreviations for the months of the year: jan, feb, . . . , dec. When an abbreviation is used, it is not enclosed in quotes (") or braces ({ }). The style defines what is actually to be typeset. Most styles typeset dec as either Dec. or December. The name of the abbreviation, such as dec, is a string of characters that starts with a letter, does not contain a space, an equal sign (=), a comma, or any of the special characters listed in Section 5.4.4. You may define your own abbreviations using the command @STRING. For example, @STRING{au = "Algebra Universalis"} A string definition can be placed anywhere in a bib file, as long as it precedes the first use of the abbreviation in an entry. The AMS supplies the mrabbrev.bib file containing the standard abbreviations for many mathematical journals. Find it at ams.org, under Reference Tools, click on MR Serials Abbreviations for BibTeX. Based on this file, you can make your own abbrev.bib file containing entries for all the journals you reference with whatever abbreviations you find easiest to remember. You should pare down the file, mrabbrev.bib, because it is too large for some systems to handle.
16.2
Using BIBTEX
like
437
If you use this scheme, the command you use to specify the bib files may look
\bibliography{abbrev,... } Section 16.2.1 explains the \bibliography command.
16.2
Using BIBTEX In Section 16.1, you learned how to create database files. The sample bib files are template.bib and sampartb.bib in the samples folder (see page 4). In this section, you learn how to use B IBTEX to process these files to create a bibliography. We illustrate the process of working with B IBTEX with the sampartb sample article. We use the amsplain style. To obtain all six examples of different styles shown on pages 422–423, just change amsplain to the appropriate style name in your document and typeset it. One B IBTEX style behaves differently. The apacite style of the American Psychological Association requires that the preamble of your document include the line \usepackage{apalike} in addition to using the style file. The package can also be modified by a large number of options.
16.2.1 Sample files Type the following two lines to replace the thebibliography environment in the sampart.tex sample document: \bibliographystyle{amsplain} \bibliography{sampartb} Save the new sample article as sampartb.tex. The first line specifies the bst file, amsplain.bst, which is part of the AMS distribution (see Section 11.6). The second line specifies the database files used. In this case there is only one, sampartb.bib. The contents of the sampartb.bib bibliographic database file are as follows: @BOOK{gM68, author = "George A. Menuhin", title = "Universal Algebra", publisher = "D.~Van Nostrand", address = "Princeton", year = 1968, }
438
Chapter 16
BIBTEX
@BOOK{fR82, author = "Ferenc R. Richardson", title = "General Lattice Theory", edition = "Expanded and Revised", language = "Russian", publisher = "Mir", address = "Moscow", year = 1982, } @ARTICLE{eM57, author = "Ernest T. Moynahan", title = "On a Problem of {M. Stone}", journal = "Acta Math. Acad. Sci. Hungar.", pages = "455460", volume = 8, year = 1957, } @ARTICLE{eM57a, author = "Ernest T. Moynahan", title = "Ideals and Congruence Relations in Lattices.~\textup{II}", journal = "Magyar Tud. Akad. Mat. Fiz. Oszt. K{\"{o}}zl.", language = "Hungarian", pages = "417434", volume = 7, year = 1957, } @PHDTHESIS{sF90, author = "SooKey Foo", title = "Lattice Constructions", school = "University of Winnebago", address = "Winnebago, MN", year = 1990, month = dec, } Type sampartb.bib or copy it from the samples folder to your work folder.
16.2
Using BIBTEX
439
16.2.2 Setup Before you start B IBTEX, make sure that everything is set up properly as described in this section. To list database entries in the bibliography, use the \cite command. Refer to Section 10.5.1 for details on how to use citations. If you want to have a reference listed in the bibliography without a citation in the text, then use the \nocite command. For example, \cite{pK57} includes the reference in the bibliography and cites the entry with label pK57, whereas \nocite{pK57} includes the reference in the bibliography but does not cite the entry. In either case, one of the bib files specified in the argument of the \bibliography command must contain an entry with the label pK57. The \nocite{*} command includes all the entries from the bibliographic databases you’ve specified. Your document must specify the bibliography style and must name the bib files to be used. For instance, the sampartb.tex sample article contains the lines \bibliographystyle{amsplain} \bibliography{sampartb} The \bibliographystyle command specifies amsplain.bst as the style and the \bibliography command specifies the database file sampartb.bib. To use several database files, separate them with commas, as in \bibliography{abbrev,gg,lattice,sampartb} where abbrev.bib contains userdefined abbreviations gg.bib contains personal articles lattice.bib contains lattice theory articles by other authors sampartb.bib contains additional references needed for sampartb.tex It is important to make sure that the bst file, the bib file(s), and the LATEX document(s) are in folders where B IBTEX can find them. If you are just starting out, you can simply copy all of them into one folder. Later, you may want to look for a more permanent solution by keeping the files abbrev.bib and lattice.bib in one “central” location, while placing sampartb.bib in the same folder as its corresponding LATEX document.
440
16.2.3
Chapter 16
BIBTEX
Four steps of BIBTEXing The following steps produce a typeset bibliography in your LATEX document. We use the sampartb.tex sample article as an example. Step 1 Check that B IBTEX, your LATEX document, and the bib files are placed in the appropriate folders. Step 2 Typeset sampartb.tex to get a fresh aux file. This step is illustrated in Figure 16.1.
LATEX
sampartb.tex the source file
sampartb.aux the aux file
typeset
Figure 16.1: Using B IBTEX, step 2.
sampartb.aux
sampartb.bbl
the aux file
bibliography file
sampartb.blg the log file
Figure 16.2: Using B IBTEX, step 3.
16.2
Using BIBTEX
441
Step 3 Run B IBTEX on the sampartb.aux file in one of the following three ways: by invoking it with the argument sampartb by starting the application and then opening sampartb.aux by running it by choosing it as a menu option of your editor or GUI front end or by clicking on an icon If B IBTEX cannot find a crucial file, for example, the bst file, it stops. The reason it stopped is shown in the log window and also written to a blg (bibliography log) file, sampartb.blg. Correct the error(s) and go back to step 2. A successful run creates a bbl (bibliography) file, sampartb.bbl, in addition to sampartb.blg. This step is illustrated in Figure 16.2. Step 4 Typeset the LATEX document sampartb.tex twice. B IBTEX uses and creates a number of files when it is run. To illustrate this process, complete the four steps using sampartb.tex. Step 1 Start fresh by deleting the aux, blg, and bbl files, if they are present. Step 2 Typeset the article sampartb.tex to get an aux file (see Figure 16.1). Notice that the log file contains warnings about missing references and a number of other lines not relevant to the current discussion. The lines in the aux file containing bibliographic information are \citation{fR82} \citation{gM68} \citation{eM57} \citation{sF90} \citation{eM57a} \bibstyle{amsplain} \bibdata{sampartb} Each \citation command in this file corresponds to a \cite or \nocite command in the article. The lines \bibliographystyle{amsplain} \bibliography{sampartb} in sampartb.tex are written as \bibstyle{amsplain} \bibdata{sampartb} in the sampartb.aux file.
442
Chapter 16
BIBTEX
Step 3 Now run B IBTEX on the sampartb.aux file (see Figure 16.2). How we do this, depends on the LATEX installation you have. In UNIX installations, you type bibtex intrarti at the command line. In newer user interfaces, B IBTEX is represented by an icon, and you drop intrarti.aux into it. In modern installations, your editor also runs B IBTEX. In WinEdt (see Section A.1.2), run B IBTEX by clicking on the B IBTEX button. In TeXShop (see Section A.2.1), change the LaTeX button to BibTeX and click on the Typeset button. Now change the BibTeX button back to LaTeX. Click on the Typeset button, and you have the typeset article with the bibliography. B IBTEX generates two new files: sampartb.blg and sampartb.bbl. Look at sampartb.blg: This is BibTeX, C Version 0.99c The toplevel auxiliary file: sampartb.aux The style: amsplain.bst Database file #1: sampartb.bib On some systems, this file may be much longer than the one I show here. At present, this blg file does not contain much important information. If there were any warnings or errors, they would be listed in this file. The sampartb.bbl file, in which B IBTEX created a thebibliography environment (see Section 10.5.1) is more interesting: \providecommand{\bysame}{\leavevmode% \hbox to3em {\hrulefill}\thinspace} \begin{thebibliography}{1} \bibitem{sF90} SooKey Foo, \emph{Lattice constructions}, Ph.D. thesis, University of Winnebago, Winnebago, MN, December 1990. \bibitem{gM68} George~A. Menuhin, \emph{Universal algebra}, D.~Van Nostrand, Princeton, 1968. \bibitem{eM57a} Ernest~T. Moynahan, \emph{Ideals and congruence relations in lattices.~\textup{II}}, Magyar Tud. Akad. Mat. Fiz. Oszt. K{\"{o}}zl. \textbf{7} (1957), 417434 (Hungarian).
16.2
Using BIBTEX
443
\bibitem{eM57} \bysame, \emph{On a problem of {M. Stone}}, Acta Math. Acad. Sci. Hungar. \textbf{8} (1957), 455460. \bibitem{fR82} Ferenc~R. Richardson, \emph{General lattice theory}, expanded and revised ed., Mir, Moscow, 1982 (Russian). \end{thebibliography} Observe that the nonbreakable spaces (ties) and the \bysame command have been provided in the author fields. Step 4 Now typeset sampartb.tex again. The typeset version now has a References section, constructed from the bbl file, but the new log file has warnings about missing entries. The new aux file contains five interesting new lines: \bibcite{sF90}{1} \bibcite{gM68}{2} \bibcite{eM57a}{3} \bibcite{eM57}{4} \bibcite{fR82}{5} These lines identify the crossreference label sF90 (see the first line shown—the symbol designates Foo’s thesis in sampartb.bib) with the number 1, and so on. Now typeset sampartb.tex again, and all the citations are correctly placed in the typeset article. Observe: 1. The crucial step 3, running the B IBTEX application, gives different error messages and obeys different rules from LATEX—see Section 16.2.4. 2. The sampartb.bbl file was created by B IBTEX. It is not changed by running LATEX.
16.2.4 BIBTEX rules and messages Rule
B IBTEX and % You cannot comment out a field with an %.
444
Chapter 16
BIBTEX
For example, the entry @ARTICLE{eM57, author = "Ernest T. Moynahan", title = "On a Problem of {M. Stone}", journal = "Acta Math. Acad. Sci. Hungar.", % pages = "455460", volume = 8, year = 1957, } causes B IBTEX to generate the error message You’re missing a field name line 23 of file sampartb.bib : : % pages = "455460", (Error may have been on previous line) I’m skipping whatever remains of this entry Warningmissing year in eM57 Warningmissing pages in eM57 (There was 1 error message) Recall that B IBTEX ignores field names it cannot recognize. So changing the field name pages, for example to pagescomment, does not give an error message. However, doing so removes a required field, so you get the warning message Warningmissing pages in eM57
Rule
B IBTEX field names Do not abbreviate field names. For instance, if you abbreviate volume to vol, as in @ARTICLE{eM57, author = "Ernest T. Moynahan", title = "On a Problem of {M. Stone}", journal = "Acta Math. Acad. Sci. Hungar.", pages = "455460", vol = 8, year = 1957, } the vol field is simply ignored. This entry is typeset as
16.2
Using BIBTEX
445
3. Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. (1957), 455–460. instead of 3. Ernest T. Moynahan, On a problem of M. Stone, Acta Math. Acad. Sci. Hungar. 8 (1957), 455–460.
Rule
B IBTEX field terminations Make sure that every field of an entry, except possibly the last, is terminated with a comma. If you drop a comma before a field, you get an error message such as I was expecting a ‘,’ or a ‘)’ line 6 of file sampartb.bib : : year = 1968, (Error may have been on previous line) I’m skipping whatever remains of this entry Warningmissing year in gM68
Rule
B IBTEX field value terminations Make sure that the field value is properly terminated. You should be careful not to drop a double quote or brace. If you drop the closing quote on line 11 of the bib file, title = "General Lattice Theory you get the error message I was expecting a ‘,’ or a ‘}’ line 12 of file sampartb.bib : edition = " : Expanded and Revised", I’m skipping whatever remains of this entry Warningmissing publisher in fR82 Warningmissing year in fR82
446
Chapter 16
BIBTEX
If, instead, you drop the opening double quote in the same line, you get the error message Warningstring name "general" is undefined line 11 of file sampartb.bib I was expecting a ‘,’ or a ‘}’ line 11 of file sampartb.bib : title = general : Lattice Theory", I’m skipping whatever remains of this entry Warningmissing title in fR82 Warningmissing publisher in fR82 Warningmissing year in fR82 (There was 1 error message) B IBTEX assumed that general was an abbreviation, since it was not preceded by a ". The obvious conclusion is that you have to be very careful about typing your bibliographic entries for B IBTEX. If you have access to special tools for maintaining your bibliographic data, use them. Otherwise, refer to the template.bib file that contains templates of oftenused entry types.
16.2.5
Submitting an article If you submit an article to a journal that provides you with a B IBTEX style file, then you can submit the article and the B IBTEX database file, pared down of course. If this is not the case, create the bbl file with amsplain.bst and copy and paste the content into the thebibliography environment in the article. Then the journal’s editor can edit the bibliography.
16.3
Concluding comments There is a lot more to B IBTEX than what has been covered in this chapter. For example, B IBTEX’s algorithm to alphabetize names is fairly complicated. Some names create additional difficulties. Where should John von Neumann be placed, under the “v”s or the “N”s? It depends on the style. How do we handle names where the first word is the family name, as in Ho Chi Minh or Gr¨atzer Gy¨orgy? Again, it depends on the style. Oren Patashnik’s BIBTEXing [52] has many helpful hints. It includes a clever hack to order entries correctly even when the style does not do so. Chapter 13 of The LATEX Companion, 2nd edition [46] has a long discussion of B IBTEX. It also contains a long list of styles. There are many tools to make B IBTEXing easier. BibDesk for the Mac is an excellent graphical B IBTEXbibliography manager. For Windows, there is BibTexMng. For UNIX , there is pybibliographer and if you are an Emacs user, there is Ebib. Written in
16.3
Concluding comments
447
Java, so available on most platforms, is JBibtexManager. There are many B IBTEX databases. The largest one may be “The Collection of Computer Science Bibliographies” with more than two million references. You can easily build your own mathematical databases with MathSciNet from the AMS. Do a search. When the result page comes up, go to the pull down menu next to Batch Download and select Citations (BibTeX). Now you can check mark the items you want by clicking on the little squares and then click on Retrieve Marked next to the pull down menu or click on Retrieve First 50. For the latter to work well, before your search, click on the Preferences button and click on the circle next to 50, so you get at most 50 items per result page. Then Retrieve First 50 retrieves them all. Finally, after many years of development, the AMS released amsrefs, the kid brother of B IBTEX, at its annual meeting in January 2002. The presentation was made by Michael Downes, who designed and coded the package. I was very excited to hear his lecture—bibliographic management was the last block needed to complete the rebuilding of LATEX. It turned out that amsrefs is not simply a B IBTEX replacement. It has a number of very important new features. 1. You can, with amsrefs, enter the marked up bibliographic entries into the document. This means that the document class of the journal publishing the paper can format your bibliography. 2. amsrefs is a LATEX package. Therefore, you do not have to learn (another) esoteric language to control the formatting of your bibliography. Developing a format for a journal is very easy. 3. The bibliographic data files are also LATEX files, so you can print them within LATEX, making it easy to maintain them. After Michael Downes passed away, David Jones took over the project, and released version 2.0 in June of 2004. Unfortunately, unlike the B IBTEX and its bbl file, amsrefs directly creates the typeset file. So if the journal you want to submit your article to does not have an amsrefs style file—and today only the AMS journals have them—then you have to manually convert the amsrefs entries into the format the journal would accept. There is no option to set in amsrefs to produce a LATEX source file for the bibliographic entries. Therefore, unless you know that you intend to submit to an AMS journal—and you know that it will accept your article for publication—you should not use amsrefs. There is a second obstacle. There is no BibDesk or BibTexMng for amsrefs. But I believe that if amsrefs overcomes the first obstacle, then the second obstacle would resolve itself fast. There is a new development which shares some of Michael Downes’ goals. It is Philipp Lehman’s biblatex package, now in version 0.6. It works with (some) B IBTEX databases and uses LATEX to format the bibliography.
CHAPTER
17 MakeIndex
Pehong Chen’s MakeIndex application, described in Pehong Chen and Michael A. Harrison’s Index preparation and processing [10], helps LATEX users create long indexes. For short indexes, you can easily do without it (see Section 10.5.2). In Section 17.1, we show you by an example how to prepare an article for indexing. We introduce formally the index commands in Section 17.2. In Section 17.3, we describe how LATEX and MakeIndex process the index entries. The rules are stated in Section 17.4. Multiple indexes are almost as easy as single indexes. They are described in Section 17.5. We conclude with glossaries in Section 17.6. Indexing is a difficult task. For an extensive discussion on how to create a useful index, consult The Chicago Manual of Style, 15th edition [11].
17.1
Preparing the document LATEX provides the theindex environment (see Section 10.5.2). Within this environment, it provides the \item, \subitem, and \subsubitem commands to typeset entries, subentries, and subsubentries, respectively, and the \indexspace command for adding vertical space between alphabetical blocks, see Figure 17.1 for an example.
450
Chapter 17
MakeIndex
The makeidx package provides the \index command for specifying the index entry at a particular point in the document, which becomes a page reference for the entry in the typeset index. Making an index entry with MakeIndex is easy. You simply place the index commands in your source file, and then let LATEX and MakeIndex do the work of gathering the entries and the page numbers for the entries, sorting them, and formatting the typeset index. There are three steps: 1. In the preamble of your LATEX document, include the line \makeindex If you do not use an AMS document class, include the two lines \usepackage{makeidx} \makeindex 2. Type the line \printindex at the point in your document where you want the index to appear, usually as part of the back matter (see Section 10.5). 3. Mark all entries in your document with \index commands. We illustrate this procedure with the intrarti.tex article, which modifies the article intrart.tex by inserting a number of index entries (both intrart.tex and intrarti.tex are in the samples folder; see page 4). We now add a dozen \index commands to intrart.tex. Command 1 Retype the line \begin{theorem} to read \begin{theorem}\index{Main Theorem} Commands 2 and 3 Type the commands \index{
[email protected]$\Pi^{*}$ construction}% \index{Main Theorem!exposition(}% after the line \section{The $\Pi^{*}$ construction}\label{S:P*}
17.1
Preparing the document Command 4 Retype the line See also Ernest~T. Moynahan~\cite{eM57a}. as follows: See also Ernest~T. \index{Moynahan, Ernest~T.}% Moynahan~\cite{eM57a}. Commands 5 to 7 Type the three index items \index{latticetextbf}% \index{lattice!distributive}% \index{lattice!distributive!complete}% before the line \begin{theorem}\label{T:P*} Command 8 Type \index{Main Theorem!exposition)} after the line hence $\Theta = \iota$. Command 9 Retype the line \bibitem{sF90} as follows: \bibitem{sF90}\index{Foo, SooKey}% Command 10 Retype the line \bibitem{gM68} as follows: \bibitem{gM68}\index{Menuhin, George~A.}%
451
452
Chapter 17
MakeIndex
Command 11 Retype the line \bibitem{eM57} as follows: \bibitem{eM57}\index{Moynahan, Ernest~T.}% Command 12 Retype the line \bibitem{eM57a} as follows: \bibitem{eM57a}\index{Moynahan, Ernest~T.}% These \index commands produce the index for the intrarti.tex article shown in Figure 17.1. Notice that although you typed 12 index commands, only 11 entries appear in the index. The last two entries for Moynahan (commands 11 and 12) occur on the same typeset page, so only one page number shows up in the index.
Figure 17.1: A simple index. The showidx package (see Section 12.3) lists all the index items of a page in a top corner on the margin. The top of the first page of the typeset intrarti.tex is shown in Figure 17.2.
17.2
453
Index commands
Figure 17.2: Using showidx.
17.2
Index commands There are a few major forms of \index commands. They are discussed in this section, illustrated by the commands shown in Section 17.1. Simple \index commands The index entry Foo, SooKey, 2 was created by command 9, \index{Foo, SooKey} This entry is an example of the simplest form of an index command: \index{entry } The entry lattice, 2 was created as command 5, \index{latticetextbf} Ignore, for the time being, the textbf part. This entry has a subentry, lattice, 2 distributive, 2 which was created by command 6,
454
Chapter 17
MakeIndex
\index{lattice!distributive} There is also a subsubentry, lattice, 2 distributive, 2 complete, 2 which was created by command 7, \index{lattice!distributive!complete} The form of the \index command for subentries is \index{entry !subentry } and for subsubentries it is \index{entry !subentry !subsubentry } Modifiers Command 5 \index{latticetextbf} produces a bold page number in the entry lattice. The command whose name follows the symbol  (in this case, the command name is textbf) is applied to the page number. For instance, if you want a large bold page number, then define the command \LargeBold as \newcommand{\LargeBoldB}[1]{\textbf{\Large #1}} and type the \index command as \index{entry LargeBold} You can also modify \index commands to indicate page ranges: Main Theorem, 1 exposition, 1–2 The latter index entry has a page range. It was created with commands 3 and 8: \index{Main Theorem!exposition(} \index{Main Theorem!exposition)}
17.2
Index commands
455
Separate an entry from its modifier with , open the page range with (, and close it with ). Modifiers can also be combined. The index commands \index{Main Theorem!exposition(textbf} \index{Main Theorem!exposition)textbf} produce a bold page range. Sorting control Observe the \index command \index{
[email protected]$\Pi^{*}$ construction} This produces the entry Π∗ construction, 1 To place this entry in the correct place in the index, use a sort key. The general form of an \index command with a sort key is \index{sortkey @entry } In this example, the sortkey is pistar. When the entries are sorted, the sortkey is used to sort the entry. A few typical examples follow: ˇ Example 1 An \index command for G.I. Zitomirski˘ ı, \index{
[email protected]\v{Z}itomirski\u{\i}, G.I.} ˇ sorts Zitomirski˘ ı with the Z entries. If you used the command \index{\v{Z}itomirski\u{\i}, G.I.} ˇ Zitomirski˘ ı would be sorted with the v’s. ˝ Example 2 An \index command for the Ormester lemma, \index{
[email protected]\H{O}rmester lemma} ˝ would sort Ormester lemma with the O entries. If you used the command \index{\H{O}rmester lemma} ˝ Ormester lemma would be sorted with the H’s.
456
Chapter 17
MakeIndex
Example 3 An \index command for truncated lattice, \index{truncated
[email protected]\emph{truncated} lattice} sorts truncated lattice with the t entries. If you use the command \index{\emph{truncated} lattice} this would sort truncated lattice with the e’s. Example 4 We want to place the symbol Truncat f , typed as \Trunc f (see Section 7.6.2) in the index, sorted as Trunc. \index{$\Trunc f$} would place Truncat f near the beginning of the index, sorted with the $ symbol. If you use the command \index{
[email protected]$\Trunc f$} this would sort Truncat f with the T’s. Sorting control and subentries If you want to place a subentry under an entry with a sort key, you must include the sort key part of the entry as well: \index{sortkey @entry !subentry } For instance, \index{
[email protected]\v{Z}itomirski\u{\i}, G.I.!education} You can also use a sort key for subentries (and subsubentries), such as \index{lattice!weakly
[email protected] \emph{weakly} distributive} or, a more complicated example, \index{
[email protected]\v{Z}itomirski\u{\i}, G.I.!elementary
[email protected]\textbf{elementary} education}
17.2
Index commands
457
Special characters Since the !, @, and  characters have special meanings within an \index command, you need to quote those characters if you want them to appear as themselves. MakeIndex uses the double quote character (") for this purpose: "!, "@, and ". Because this usage makes the double quote a special character itself, it also has to be quoted if you need to use it in an \index command: "". Example 1 To produce the entry Start here!, type the \index command as \index{Start here"!} Example 2 To produce the entry @ symbol, type the \index command as \index{"@ symbol} Example 3 To produce the entry A, type the \index command as \index{"A"@$"A"$}
Crossreferences It is easy to make a crossreference to another index entry. For instance, to list distributive lattice by crossreferencing it to lattice, distributive, the command is \index{distributive latticeseeonly{lattice, distributive}} which produces the entry distributive lattice, see lattice, distributive For nonAMS document classes, seeonly should be see. A command of this form can be placed anywhere in the document.
Tip
Put all crossreferencing \index commands in one place in your document, so they are easy to keep track of.
458
Chapter 17
MakeIndex
Placement of \index commands The principle is simple.
Rule
Placement of \index commands An \index command should: 1. Reference the correct page 2. Not introduce unwanted space into the typeset document For example, you should avoid placing \index commands as shown here: Let $L$ be a distributive lattice \index{lattice} \index{distributive lattice} that is strongly complete. This placement may result in unwanted extra space following the word lattice: Let L be a distributive lattice that is strongly complete. Note the placement of the \index commands in Section 17.1. In each case I have placed them as close to the referenced item as I could. If you place an index entry on a separate line, use % to comment out unwanted spaces including the endofline character (see Section 5.5.1), as in Let $L$ be a distributive lattice \index{lattice}% \index{distributive lattice}% that is strongly complete. Read also Section 18.5 on page breaks and index entries. Listing the forms of the \index command We have discussed the following forms: \index{entry } \index{entry !subentry } \index{entry !subentry !subsubentry } \index{entry modifier } \index{entry open/close modifier } \index{sortkey @entry }
17.3
Processing the index entries
459
\index{sortkey @entry !subentry } \index{sortkey @entry !subsortkey @subentry } Of course, more combinations are possible; the following may be the longest form: \index{sortkey @entry !subsortkey @subentry % !subsubsortkey @subsubentry open/close modifier }
17.3
Processing the index entries Once you are satisfied with the \index commands, the index is ready to be created. Step 1 Typeset intrarti.tex (see Figure 17.3). Step 2 Run the MakeIndex application on intrarti.idx (see Figure 17.4). Step 3 Typeset intrarti.tex again. You find the index on page 3 of the typeset document. Let us look at this process in detail. In step 1 (see Figure 17.3), LATEX creates the intrarti.idx file: \indexentry{Main Theorem}{1} \indexentry{
[email protected]$\Pi^{*}$ construction}{1} \indexentry{Main Theorem!exposition(}{1} \indexentry{Moynahan, Ernest~T.}{1} \indexentry{latticetextbf}{1} \indexentry{lattice!distributive}{1} \indexentry{lattice!distributive!complete}{1} \indexentry{Main Theorem!exposition)}{2} \indexentry{Foo, SooKey}{2} \indexentry{Menuhin, George~A.}{2} \indexentry{Moynahan, Ernest~T.}{2} \indexentry{Moynahan, Ernest~T.}{2} In step 2 (see Figure 17.4), MakeIndex processes intrarti.idx and creates the index file intrarti.ind, which contains a theindex environment with all the index entries: \begin{theindex} \item Foo, SooKey, 2 \indexspace
460
Chapter 17
intrarti.tex
LATEX
the source file
typeset
intrarti.idx index entry listing file
Figure 17.3: Using MakeIndex, step 1.
intrarti.idx
MakeIndex
intrarti.ind index file
index entry listing file
intrarti.ilg the log file
Figure 17.4: Using MakeIndex, step 2. \item lattice, \textbf{1} \subitem distributive, 1 \subsubitem complete, 1 \indexspace \item Main Theorem, 1 \subitem exposition, 12 \item Menuhin, George~A., 2
MakeIndex
17.3
Processing the index entries
461
\item Moynahan, Ernest~T., 1, 2 \indexspace \item $\Pi^{*}$ construction, 1 \end{theindex} The \printindex command reads intrarti.ind during the next typesetting cycle. MakeIndex also produces the index log file intrarti.ilg: This is makeindex, version 2.14 [02Oct2002] (kpathsea + Thai support). Scanning input file intrarti.idx....done (12 entries accepted, 0 rejected). Sorting entries....done (43 comparisons). Generating output file intrarti.ind....done (22 lines written, 0 warnings). Output written in intrarti.ind. Transcript written in intrarti.ilg. It is important to understand that in step 1, LATEX does not process the index entries, it simply writes the arguments of the \index commands in the source file to the idx file as arguments of \indexentry commands verbatim (that is, with no change). MakeIndex then processes the idx file by removing the double quote marks for the special characters, sorting the entries, and collating the page numbers. The resulting ind file is a normal LATEX source file (you can edit it, if necessary) that is included in the original document by the \printindex command the next time you run LATEX. In Step 2, we run the MakeIndex application on intrarti.idx. How we do this depends on the LATEX installation you have. In UNIX installations, you type MakeIndex intrarti at the command line. In newer user interfaces, MakeIndex is represented by an icon, and you drop intrarti.idx into it. In modern installations, your editor also runs MakeIndex. In WinEdt (see Section A.1.2), run MakeIndex by clicking on the MakeIndex button. In TeXShop (see Section A.2.1), change the LaTeX button to MakeIndex with Typeset>MakeIndex and click on the Typeset button. Now change the MakeIndex button back to LaTeX with Typeset>LaTeX. Click on the Typeset button, and you have the typeset article with the index.
462
17.4
Chapter 17
MakeIndex
Rules There are some simple rules to keep in mind when entering index items.
Rule
Spaces in \index Do not leave unnecessary spaces in the argument of an \index command. \index{item}, \index{ item}, and \index{item } produces three different entries. There are options that instruct MakeIndex to ignore such spaces, but you are better off typing the \index commands correctly in the first place.
Rule
Spacing rules for MakeIndex LATEX’s text spacing rules (Section 5.2.1) do not apply. MakeIndex does not follow these rules when it sorts the index items. While LATEX ignores spaces, MakeIndex does not.
Rule
Sort keys In \index{sortkey @item }, the sortkey is both space and case sensitive. For instance, \index{
[email protected]$\alpha$} \index{
[email protected]$\alpha$} \index{
[email protected]$\alpha$} represent three different items.
Rule
Braces In every entry, the braces must be balanced. Normally, balancing braces is not a problem. The braces within a math formula or a TEX expression should always be balanced. However, the \index command that creates the entry for { with the sort key leftbrace cannot be typed as \index{@\{} because LATEX would give the error message
17.5
Multiple indexes
463
Runaway argument? {
[email protected]\{} ! Paragraph ended before \@wrindex was complete. There are many ways to correct this \index command. Perhaps the simplest is to define \newcommand{\printleftbrace}{\{} and rewrite the \index command \index{@\printleftbrace} This produces the entry {, 1 There is, of course, a lot more to MakeIndex than what we have discussed in this short introduction, but what we have covered here should do for most documents. See Pehong Chen and Michael A. Harrison’s Index preparation and processing [10] for more detail. Chapter 12 of The LATEX Companion, 2nd edition [46] covers MakeIndex in great detail, including the customization of indexes.
17.5
Multiple indexes To split your index file into two parts, the first with all names and the second with the other entries, include the following three lines in the preamble: \usepackage{amsmidx} \makeindex{names} \makeindex{others} For the package amsmidx, see Section 11.6. Now enter each index command in the form \index{names}{} or \index{others}{} So you would enter \index{Moynahan, Ernest~T.} as \index{names}{Moynahan, Ernest~T.}
464
Chapter 17
MakeIndex
and \index{Main Theorem} as \index{others}{Main Theorem} Where you want the two indexes to appear in the typeset article, enter the commands \Printindex{names}{Contributors} \Printindex{others}{Index} Of course, instead of Contributors and Index you may enter any titles for the two indexes.
17.6
Glossary Using the glossary commands is very similar to using the corresponding index commands. Instead of the \index and \makeindex commands, use the \glossary and \makeglossary commands, respectively. Glossary entries are written in the glo file, which corresponds to the idx file. LATEX gives you no further assistance in making a glossary file. There is no \printglossary command, theglossary environment, or MakeGlossary application. There is, however, Thomas Henlich’s makeglos package (see Section E.1).
17.7
Concluding comments Style files are available in a very limited form also for MakeIndex. Google makeindex style for an uptodate listing. Indexing is a very complex process, so it is not surprising that there are many index packages available. CTAN lists a number of them in the index directory. The best known is xindy, described in detail in The LATEX Companion, 2nd edition [46].
CHAPTER
18 Books in LATEX
Since the introduction of LATEX, the visual quality of articles published in mathematical journals has improved dramatically. Unfortunately, the same cannot be said of books published using LATEX. A record number of very ugly books have appeared. It is easy to understand why. While amsart has been designed to produce highquality printed output, the standard book document classes do not produce attractive books without additional work. LATEX provides the book and the amsbook document classes to serve as foundations for welldesigned books. Better quality books have to use document classes designed by professionals. We provide some sample pages from a book using Springer’s svmono.cls document class. We briefly discuss logical and visual design in Section 4.3.2. So this chapter is not about how to produce a finished book using LATEX. Our goal is much more modest, how to prepare a book manuscript for your publisher. In Section 18.1 we describe the book document classes book and amsbook. The table of contents and lists of figures and tables are discussed in Section 18.2. Typesetting a book involves dozens of files. Section 18.3.3 gives some tips on how to organize them. Section 18.4 covers logical design. Section 18.5 deals with the final preparation of your edited manuscript for your publisher. Finally, Section 18.6
466
Chapter 18
Books in LATEX
suggests a few more things to do if you typeset your book yourself.
18.1
Book document classes In this section, we briefly discuss the way in which book and amsbook, the two standard book document classes, differ from the corresponding article document classes. We also show a few pages from a book that was typeset with Springer Verlag’s document class for monographs: svmono.cls.
18.1.1
Sectioning Book document classes have chapters, invoked with the \chapter command and parts, invoked with \part. The \part command is generally used to group chapters in longer documents, for instance in this book. Parts have no effect on the numbering of chapters, sections, and so on, so Chapter 1 of Part I is not numbered as I.1 but as 1. Both \chapter and \part take a title as an argument, but the \chapter command also has an optional argument: \chapter[short_title ]{title } The optional short title argument is used in the running head. You may need to protect any fragile commands in title and short title with the \protect command (see Section 5.3.3). Here is the whole hierarchy: \part \chapter \section \subsection \subsubsection \paragraph \subparagraph Book document classes, as a rule, do not number subsubsections or any of the sectioning divisions below that level. Equations in chapters By default, equations are numbered from 1 within chapters. So in Chapter 1 as well as in Chapter 3, the equations are numbered (1), (2), and so forth. If you have the \numberwithin{equation}{chapter} command in the preamble, then equations in Chapter 2 are numbered as (2.1), (2.2), and so on.
18.1
Book document classes
467
18.1.2 Division of the body The book document classes formalize the division of the body into three parts. Front matter The material that appears in the front of the document, including the title pages (normally four), table of contents, preface, introduction, and so on. LATEX numbers these pages using roman numerals. The front matter is introduced with the \frontmatter command. Main matter The main part of the book, including the appendices if any. Page numbering starts from 1 using arabic numerals. The main matter is introduced with the \mainmatter command. Back matter Material that appears in the back of the book, including the bibliography, index, and various other sections, such as the colophon, afterword, and so on. The back matter is introduced with the \backmatter command. For the book document class—and the document classes built on it—in the front and back matter, the \chapter command does not produce a chapter number but the title is listed in the table of contents. So you can start your introduction with \chapter{Introduction} Within such a chapter, you should use the *ed forms of the sectioning commands \section, \subsection, and so on, otherwise you have sections with numbers such as 0.1. In the main matter, the \appendix command marks the beginning of the appendices. Each subsequent chapter becomes a new appendix. For example, \appendix \chapter{A proof of the Main Theorem}\label{A:Mainproof} produces an appendix with the given title. Note that appendices may be labeled and crossreferenced. In Appendix A, sections are numbered A.1, A.2, and so on, subsections in A.1 are numbered A.1.1, A.1.2, and so on. The precise form these numbers take depends, of course, on the document class, packages, and userspecific changes (see Section 15.5.1). See Section 18.3.1 for a detailed example. For the amsbook document class—and the document classes built on it—the \chapter and \chapter* commands always produce a title listed in the table of contents. The following two questions are frequently asked: My book has only one appendix. How can I get it to be called just “Appendix”, not “Appendix A”?
468
Chapter 18
Books in LATEX
The single appendix in my book is being labeled “Appendix A”. How can I change this to just “Appendix”? This appendix has a title, so the answer to the preceding question doesn’t apply. These questions are answered in the author FAQ of the AMS, go to http://www.ams.org/authors/authorfaq.html
18.1.3
Document class options The options and defaults for the book document classes are the same as those of other document classes (see Sections 11.5 and 12.1.2) with a few exceptions. Twosided printing Options:
twoside oneside
default
The twoside option formats the output for printing on both sides of a page. Titlepage Options:
titlepage notitlepage
default
The titlepage option creates a separate title page. The notitlepage option creates no separate pages. Chapter start openright default openany A chapter always starts on a new page. The book document class—and the document classes built on it—use the option openright to start each chapter on an odd page, while the option openany starts each chapter on the first available new page. If you use the default option, end each chapter with the command Options:
\cleardoublepage Then if a chapter ends on an odd page, a blank page is added with no header or page number. The \cleardoublepage command is correctly coded if you use amsbook. Otherwise, use the package cleardoublepage.sty (in the samples folder). The amsbook document class—and the document classes built on it—automatically clears to a righthand page and leaves a totally blank page if needed.
18.1
Book document classes
469
18.1.4 Title pages The book document class supports the commands: \title, \author, \date, and \maketitle (see Section 12.1.1). The amsbook document class supports the same commands as amsart (see Section 11.2). You can design your own title page within the titlepage environment, which does not require the use of the \maketitle command. Title pages for books, of course, should be created by a book designer for the publisher.
18.1.5 Springer’s document class for monographs We choose svmono.cls, Springer’s document class for monographs, to demonstrate the sophisticated appearance of a book typeset with a professionally designed document class. On the next three pages, I display the title page, the first Table of Contents page, and the first page of Chapter 1 of Claudio Procesi’s book on Lie groups.
470
Chapter 18
Books in LATEX
Claudio Procesi
Universit`a di Roma La Sapienza
Lie Groups
An Approach through Invariants and Representations
18.1
471
Book document classes
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Conventional Notations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxiii 1
General Methods and Ideas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Groups and Their Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Symmetric Group . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Group Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Orbits, Invariants and Equivariant Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Orbits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Stabilizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Invariants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Basic Constructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.5 Permutation Representations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Invariant Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7 Commuting Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Linear Actions, Groups of Automorphisms, Commuting Groups . . . . . . 3.1 Linear Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 The Group Algebra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Actions on Polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4 Invariant Polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Commuting Linear Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 1 1 2 4 4 5 7 7 8 10 11 11 11 13 15 16 17
2
Symmetric Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Symmetric Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 Elementary Symmetric Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Symmetric Polynomials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Resultant, Discriminant, B´ezoutiant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Polynomials and Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19 19 19 21 22 22
472
Chapter 18
Books in LATEX
1 General Methods and Ideas
Summary. In this chapter we will develop the formal language and some general methods and theorems. To some extent the reader is advised not to read it too systematically since most of the interesting examples will appear only in the next chapters. The exposition here is quite far from the classical point of view since we are forced to establish the language in a rather thin general setting. Hopefully this will be repaid in the chapters in which we will treat the interesting results of Invariant Theory.
1 Groups and Their Actions 1.1 Symmetric Group In our treatment groups will always appear as transformation groups, the main point being that, given a set X , the set of all bijective mappings of X into X is a group under composition. We will denote this group S(X ) and call it the symmetric group of X . In practice, the full symmetric group is used only for X a finite set. In this case it is usually more convenient to identify X with the discrete interval {1, . . . , n} formed by the first n integers (for a given value of n). The corresponding symmetric group has n! elements and it is denoted by Sn . Its elements are called permutations. In general, the groups which appear are subgroups of the full symmetric group, defined by special properties of the set X arising from some extra structure (such as from a topology or the structure of a linear space, etc.). The groups of interest to us will usually be symmetry groups of the structure under consideration. To illustrate this concept we start with a definition: Definition. A partition of a set X is a family of nonempty disjoint subsets Ai such that X = ∪i Ai . A partition of a number n is a (nonincreasing) sequence of positive numbers: m 1 ≥ m 2 ≥ · · · ≥ m k > 0 with
k j=1
m j = n.
18.2
Tables of contents, lists of tables and figures
18.2
473
Tables of contents, lists of tables and figures A long document, as a rule, has a table of contents. It may also include a list of figures and a list of tables.
18.2.1 Tables of contents What goes into the table of contents? For the amsbook document class—and the document classes built on it—all titles, not the short titles, of the sectioning commands, whether *ed or not, subject only to the value of the tocdepth counter, as described in the last subsection of Section 15.5.1. For instance, if tocdepth is set to 2, the default, then the titles of chapters, sections, and subsections are included in the table of contents, and subsubsections are excluded. This leaves us with the problem, what do we do if the title is too long? You cannot break the line with \\, because this would the appear in table of contents. The AMS coded the following solution: enter the line break in the form \except{toc}{\linebreak} For the book document class—and the document classes built on it—the title or optional argument of the sectioning commands, subject to the value of the tocdepth counter, with the following exceptions: In Section 10.4.1 we discuss the *ed versions of sectioning commands. They are excluded from the table of contents. If the sectioning command has a short title, then it is the short title that is utilized. The example in Section 11.2 shows why this is important. If you have \\ in the title, you must have a short title without it, otherwise the linebreak would show up in the running head and the table of contents. When you typeset your document with a table of contents, LATEX creates a file with the toc extension. The next time the document is typeset, the toc file is typeset too and included in your typeset document at the point where the command \tableofcontents appears in the source file, normally in the front matter. If your source file is named myart.tex, the toc file is named myart.toc. This file lists all the sectioning units as well as their titles and page numbers. If you already have a toc file, the \tableofcontents command typesets a table of contents using the previously created toc file and creates a new toc file. LATEX adds a line to the table of contents, formatted like a section title, if you include the command \addcontentsline{toc}{section}{text to be added } in your source file. There are three arguments:
474
Chapter 18
Books in LATEX
1. The first argument informs LATEX that a line, the third argument, should be added to the toc file. 2. The second argument specifies how the line should be formatted in the table of contents. In our example, the second argument is section, so the line is formatted as a section title in the table of contents. The second argument must be the name of a sectioning command. 3. The third argument is the text to be added. You can add an unformatted line to the table of contents with the command \addtocontents{toc}{text to be added } Such a command can also be used to add vertical spaces into the table of contents. For instance, if you want to add some vertical space before a part, you should insert the following line before the sectioning command for the part: \addtocontents{toc}{\protect\vspace{10pt}} The toc file is easy to read. The following are typical lines from the table of contents file for a document using the book document class: \contentsline{section}{\numberline {54.}Top matter}{119} \contentsline{subsection}{\numberline {54.1.} Article info}{119} \contentsline {subsection}{\numberline {54.2.} Author info}{121} Section 15.5.1 explains how you can specify which levels of sectioning appear in the table of contents. Section 2.3 of The LATEX Companion, 2nd edition [46] lists the style parameters for the table of contents. It also shows you how to define new toclike files and use multiple tables of contents in a single document, for instance, adding a mini table of contents for each chapter.
Tip
You may have to typeset the document three times to create the table of contents and set the numbering of the rest of the document right. 1. The first typesetting creates the toc file. 2. The second inserts the table of contents with the old page numbers into the typeset document, rerecords in the aux file the page numbers, which may have changed as a result of the insertion, and crossreferences in the aux file, and generates a new toc file with the correct page numbers.
18.2
Tables of contents, lists of tables and figures
475
3. The third typesetting uses these new aux and toc files to typeset the document correctly and creates a new toc file. Fragile commands in a movable argument, such as a section (short) title, must be \protected (see Section 5.3.3). Here is a simple example using the table of contents. If the document contains the \section command \section{The function \( f(x^{2}) \)} the section title is stored in the toc file as \contentsline {section}{\numberline {1}The function\relax $ f(x^{2}) \relax \GenericError { } {LaTeX Error: Bad math environment delimiter}{Your command was ignored.\MessageBreak Type I to replace it with another command,\MessageBreak or to continue without it.}}{1} and the log file contains the message ! LaTeX Error: Bad math environment delimiter. ... l.1 ...continue without it.}}{1} Error messages usually refer to a line in the source file, but in this case the error message refers to a line in the toc file. The correct form for this section title is \section{The function \protect\( f(x^{2}) \protect\)} or, even simpler, \section{The function $f(x^{2})$} Note that this example is merely an illustration of unprotected fragile commands in movable arguments. As a rule, avoid using formulas in (sectioning) titles.
18.2.2 Lists of tables and figures If you place a \listoftables command in the document, LATEX stores information for the list of tables in a lot file. The list of tables is inserted into the body of your document at the point where the command appears, normally in the front matter, following the table of contents. A list of figures, similar to a list of tables, can be compiled with the command \listoffigures. This command creates an auxiliary file with the extension lof.
476
Chapter 18
Books in LATEX
An optional argument of the \caption commands in your tables and figures can replace the argument in the list of tables and figures. Typically, the optional argument is used to specify a shorter caption for the list of tables or list of figures. There are other uses. For instance, you may notice that, as a rule, captions should be terminated by periods. If in the list of tables or list of figures, your book style fills the space between the text and the page number with dots, the extra period looks bad. This problem goes away if you use the following form of the \caption command: \caption[title ]{title. } There are analogs of the table of contents commands for use with tables and figures. The command \addtocontents{lot}{line_to_add } adds a line to the list of tables or to the list of figures with the first argument lof.
18.2.3
Exercises For exercises, amsbook provides the xcb environment. It is used for a series of exercises at the end of a section or chapter. The argument of the environment specifies the phrase (such as Exercises) to begin the list: \begin{xcb}{Exercises} \begin{enumerate} \item A finite lattice $L$ is modular if{f} it does not contain a pentagon.\label{E:pent} \item Can the numbers of covering pairs in\label{E:incr} Exercise~\ref{E:pent} be increased?\label{E:incr} \end{enumerate} \end{xcb} which typesets as
Exercises (1) A finite lattice L is modular iff it does not contain a pentagon. (2) Can the numbers of covering pairs in Exercise 1 be increased?
18.3
Organizing the files for a book An article is typically one tex file and maybe some EPS and/or PDF files for the illustrations. On the other hand, a book, like this one, is composed of hundreds of files. In this section, I describe how the files for this book are organized.
18.3
Organizing the files for a book
477
There are three commands that help with the organization: \include \includeonly \graphicspath We discuss these commands in this section.
18.3.1 The folders and the master document All the files for this book are in a folder MiL4 and in this folder the most important document is MiL4.tex, the master document. The master document, MiL4.tex reads, in a somewhat simplified form, as follows: %MiL4 master document \documentclass[leqno]{book} \usepackage{MiL4} \usepackage{makeidx} \makeindex \usepackage{cleardoublepage} \includeonly{ %frontmatter, %intro, %Chapter1,% terminology ... Chapter18,% books %appA,% install ... } \begin{document} \frontmatter \include{frontmatter} \tableofcontents \listoftables\listoffigures \include{intro}%Intro \mainmatter \include{Chapter1}%Setting up ... \include{Chapter18}% writing books \appendix \include{AppendixA}%install ... \backmatter \printindex \end{document}
478
Chapter 18
Books in LATEX
Some parts of the master file deserve comment, for example, the third line, \usepackage{MiL4} loads the command file MiL4.sty which contains all the commands defined for the book and the code for the book style. Since the book style is based on book.cls, in line 5, we load the makeidx package and print the index with \printindex (see Section 17.1). Line 7 states \usepackage{cleardoublepage} This creates blank pages after chapters that end on an odd page number (see Section 18.1.3). For the three ...matter commands, see Section 18.1.2.
18.3.2
Inclusion and selective inclusion This book is pieced together by the \include commands in the master document. For example, \include{Chapter18} inserts the contents of the file Chapter18.tex, starting on a new page, as though its contents had been typed at that place in the document. The master document for this book has 29 \include commands.
Rule
File termination Terminate every file you \include with an \endinput command. If you terminate an \includeed file with \end{document}, LATEX gives a warning such as: (\end occurred when \iftrue on line 6 was incomplete) (\end occurred when \ifnum on line 6 was incomplete) If you use \include commands in the master file, as in the example in Section 18.3.1, then you can use the \includeonly command for selective inclusion. The lines of the \includeonly command parallel the \include commands. Block comment all the lines of the argument of the \includeonly command, and uncomment the chapter you are working on. In the example above, I am working on this chapter. The argument of the \includeonly command is a list of files separated by commas. If you want to typeset the whole book, uncomment all the lines.
18.4
Logical design
479
18.3.3 Organizing your files The MiL4 folder, containing the files of this book, contains the master document, MiL4.tex, the command file MiL4.sty, and all the tex files listed in the master document, that is, the chapters, the frontmatter, the introduction, the appendices, and of course, all the auxiliary files that LATEX creates. This book contains about 300 illustrations in a subfolder Graphics of the folder MiL4.tex. We have to tell LATEX to look for the illustrations in this folder. We do this with the command \graphicspath{{Graphics\}} in the preamble. If you have two folders, Illustr1 and Illustr2 for illustrations, the \includegraphics command takes the form \graphicspath{{Illustr1\}{Illustr2\}} Even if you have more than one folder for the illustrations, you must make sure that each graphics file has a unique name. We place the \graphicspath command in MiL4.sty. In the above commands, \ is appropriate for Mac and UNIX computers. For a PC, use / instead. If you submit a dvi file, you cannot use the \graphicspath command.
18.4
Logical design The discussion of logical and visual design in Section 4.3.2 applies to books even more than to articles. Since books are long and complex documents, errors in the logical design are much harder to correct. Let us review some commonsense rules.
Rule 1
Stick with the sectioning commands provided by the document class. Define the nonstandard structures you wish to use as environments. Here is an example which is obviously bad: \vspace{18pt} \noindent \textbf{Theorem 1.1.} \textit{This is bad.} \vspace{18pt}
480
Chapter 18
Books in LATEX
And a good way to achieve the same result: \begin{theorem}\label{T:Goodtheorem} This is a good theorem. \end{theorem} The bad example creates a number of difficulties. You have to number the theorems yourself. Adding, deleting, and rearranging theorems becomes difficult and updating crossreferences is even harder. It is difficult to keep such constructs consistent. If the publisher decides to increase the white space before and after the theorems to 20 points, finding and changing all the appropriate commands becomes a tedious and error prone task.
Rule 2
Define frequently used constructs as commands. Rather than \textbf{Warning! Do not exceed this amount!} define \newcommand{\important}[1]{\textbf{#1}} and type your warnings as \important{Warning! Do not exceed this amount!} You or your editor can then change all the warnings to a different style with ease.
Rule 3
Avoid text style commands. If you use small caps for acronyms, do not type \textsc{ibm} but rather define \newcommand{\ibm}{\textsc{ibm}} and then
18.4
481
Logical design \ibm or more generally \newcommand{\acronym}[1]{\textsc{#1}} and then \acronym{ibm}
Rule 4
Avoid white space commands. Occasionally, you may feel that there should be some white space separating two paragraphs, so you do the following: paragraph 1 \medskip paragraph 2 It would be better to define a new command, say \separate, as \newcommand{\separate}{\medskip} and type the previous example as paragraph 1 \separate paragraph 2 Now such white space can be adjusted throughout the entire document by simply redefining one command. Note that redefining \medskip itself may have unintended side effects: Many environments depend on LATEX’s definition of \medskip. You may have used \medskip in other situations as well. Here is a short list of commands you should avoid: \bigskip \break \eject
\hfil \hfill \hglue
\hspace \kern \medskip
\parskip \smallskip \vfil
\vfill \vglue \vskip
\vspace
482
18.5
Chapter 18
Books in LATEX
Final preparations for the publisher Throughout this book, there are a number of “don’ts”. Most are practices you should avoid while writing articles. When writing a book, it is even more important not to violate these rules. When the editors, including the copy editor, are finished with your manuscript and you have the document class designed for the book, then you can start on the final preparations.
Step 1
Eliminate all TEX commands. TEX commands, that is, Plain TEX commands that are not part of LATEX (not listed as LATEX commands in the index of this book) may interfere with LATEX in unexpected ways. They may also cause problems with the style file that is created for your book.
Step 2
Collect all your userdefined commands and environments together in one place, preferably in a separate command file (see Section 15.3).
Step 3
Make sure that userdefined commands for notations and userdefined environments for structures are used consistently throughout your document. This book uses the command \doc for document names, so intrart is typed as \doc{intrart}. Of course, \texttt{intrart} gives the same result, but if you intermix \doc{intrart} and \texttt{intrart} commands, you lose the ability to easily change the way document names are displayed.
Step 4
Watch out for vertical white space adding up. Do not directly follow one displayed math environment with another. Multiple adjacent lines of displayed mathematics should all be in the same environment. If your style file uses interparagraph spacing, avoid beginning paragraphs with displayed math. For instance, \[ \]
x=y
18.5
Final preparations for the publisher \[
483
x=z
\] is wrong. Use, instead, an align or gather environment.
Step 5
If possible, do not place “tall” mathematical formulas inline. All formulas that might change the interline spacing, as a rule, should be displayed. You can find examples on pages 23 and 370. Here is one more example, double hat ˆ ˆ accents used inline: A.
Step 6
Read the log file. Watch for linetoowide warnings (see Section 2.3). Check for font substitutions (see Section 5.6.7). If you find lines that are too wide: Fix wide lines by rewording the sentence or adding optional hyphens (see Section 5.4.9). Break displayed formulas so that they fit comfortably within the line. Adobe Acrobat Professional has a preflight utility in the Advanced menu. It will check whether the PDF version of your typeset document has all the fonts it requires.
Step 7
Do not assume that gray boxes or color illustrations appear when published exactly the way that they look on your monitor or printer. Color work requires calibration of monitors and printers. It is often best left to the experts at the publisher.
Step 8
Do not assume that the application that created your EPS files (see Section 10.4.3) can create highquality EPS files. Many applications can create EPS files or convert files to EPS format. Very few do it right. Ask your publisher what applications they recommend.
484
Chapter 18
Books in LATEX
Font substitutions can also cause problems: A font that was used in typesetting the document may not be the font you intended. Missing fonts are substituted and the substitute fonts are rarely satisfactory. A special trap: Your publisher may have more, or maybe fewer, fonts than you do! As a result, the font substitutions on your publisher’s system may be different from those on yours. Make sure that the fonts you use are not substituted.
18.6
If you create the PDF file for your book Many publishers take your manuscript, prepared as described in Section 18.5, and guide it through the final steps for printing. Some books, however, are prepared by the authors for printing using a custom document class for books and submitted to the publisher as PDF files. If your book falls into this category, there are a few more things you should do before you create the final PDF file for your book.
Adjust the pages Make sure that you are satisfied with the way the document is broken into pages by LATEX and with the placement of the figure and table environments (see Section 10.4.3). If necessary, you should make lastminute changes to adjust page breaks. You may find the \enlargethispage command (see Section 5.7.3) very helpful at this stage. Just be sure to apply it on both facing pages. To ensure that Page numbers in the index are correct \pageref references (see Section 10.4.2) are correct Marginal comments (see Section 5.9.4) are properly placed Tables and figures are properly placed insert page breaks where necessary. Where pages break, add the three commands \linebreak, \pagebreak, and \noindent. Here is an example. The bottom of page 3 and the top of page 4 of my book General Lattice Theory [28] are shown in Figure 18.1. Now let us assume that we have to manually do this page break because some index items attached to this paragraph generate incorrect page numbers. The paragraph split by the page break is In other words, lattice theory singles out a of poset for detailed investigation. To make definition worthwhile, it must be shown that of posets is a very useful class, that there
special type such a this class are many
18.6
If you create the PDF file for your book
485
such posets in various branches of mathematics (analysis, topology, logic, algebra, geometry, and so on), and that a general study of these posets will lead to a better understanding of the behavior of the examples. This was done in the first edition of G.~Birkhoff’s \emph{Lattice Theory} \cite{gB40}. As we go along, we shall see many examples, most of them in the exercises. For a general survey of lattices in mathematics, see G.~Birkhoff \cite{gB67} and H.~H.~Crapo and G.C.~Rota \cite{CR70}. When typesetting this paragraph, LATEX inserts a page break following This was done in the first edition of G.~Birkhoff’s. So we edit four lines as follows: understanding of the behavior of the examples. This was done in the first edition of G.~Birkhoff’s \linebreak \pagebreak \noindent \emph{Lattice Theory} \cite{gB40}. As we go along, we shall see many examples, most of them in the This change does not affect the appearance of the typeset page, but now pages 3 and 4
In other words, lattice theory singles out a special type of poset for detailed investigation. To make such a definition worthwhile, it must be shown that this class of posets is a very useful class, that there are many such posets in various branches of mathematics (analysis, topology, logic, algebra, geometry, and so on), and that a general study of these posets will lead to a better understanding of the behavior of the examples. This was done in the first edition of G. Birkhoff’s
4
I. First Concepts
Lattice Theory [1940]. As we go along, we shall see many examples, most of them in the exercises. For a general survey of lattices in mathematics, see G. Birkhoff [1967] and H. H. Crapo and G.C. Rota [1970]. Figure 18.1: A page break.
486
Chapter 18
Books in LATEX
are separated by a \pagebreak. Make sure that any \index or \label commands are moved to the appropriate half of the paragraph. Now all index commands generate the correct page numbers. Of course, if the page break is between paragraphs, only the \pagebreak command is needed. If the break occurs in the middle of a word, use \\linebreak to add a hyphen. This method works about 95 percent of the time. Occasionally, you have to drop either the \linebreak or the \pagebreak command.
Check for missing fonts and other defects Open the PDF file of your book in Adobe Reader (or even better, in Adobe Acrobat Pro). Under File, go to Properties... and click on the Fonts tab. You will find a long list of fonts. Each one should be marked Embedded Subset. If all your fonts are embedded, you are in good shape. Adobe Acrobat Pro has an excellent set of utilities to check whether your PDF file is ready for printing. You find them under Advanced>Preflight.... In the Preflight window, choose Digital printing (B/W)—unless your book will print in color, in which case choose Digital printing (color). Click on Execute. Adobe Acrobat Pro will correct all the mistakes it finds in the file and presents a detailed report. Adobe Acrobat Pro also comes to the rescue if some fonts are not embedded. In the Preflight window, expand the PDF analysis group and select List text using nonembedded fonts and click on Execute. The report will list all pages with fonts missing and if you select a page, Snap view will show you the trouble spot. In the help system of Adobe Acrobat Pro, search for Customize Adobe PDF settings and Embed fonts using the TouchUp Text tool for detailed instructions on how to embed the missing fonts.
Other adjustments Move the figure and table environments (see Section 10.4.3) physically close to where they appear in the typeset version, and change the optional argument of the figure and table environments to !h. Balance the white space on each page as necessary. Generate the index only after the page breaks are fixed.
Polish the auxiliary files Typeset the document one last time and then place the \nofiles command in the preamble (see Section D.3.4) to make sure that the auxiliary files are not overwritten.
18.6
If you create the PDF file for your book
487
Normally, you should not have to edit the table of contents (toc) file or the lot and lof files (see Section 18.2) and your style file should take care of the formatting. Sometimes, however, an unfortunate page break makes editing necessary. In an appropriate place, you may want to add to the text the command \addtocontents{toc}{\pagebreak} to avoid such edits. Create the index (ind) file from the new aux file, as described in Section 17.3. The Chicago Manual of Style, 15th edition [11] has a section on bad breaks, remedies, and Continued lines in the index. Break the ind file into pages. To minimize bad breaks, use the \enlargethispage command where necessary (see Section 5.7.3). Add any Continued entries.
APPENDIX
A Installation
In case you do not already have a LATEX installation, in Sections A.1 and A.2, we describe how to install LATEX on your computer, a PC or a Mac. The installation is much easier if you obtain TEX Live 2007 (or later) from the TEX Users Group, TUG (see Section E.2). It contains both the TEX implementations we discuss. No installation is given for UNIX computers. The attraction of UNIX to its users is the incredibly large number of options, from the UNIX dialect, to the shell, the editor, and so on. A typical UNIX user downloads the code and compiles the system. This is obviously beyond the scope of this book. Nevertheless, TEX Live 2007 (or later) from the TEX Users Group supplies the compiled (binaries) of LATEX for a number of UNIX variants. First read Chapter 1, so that in this Appendix you recognize the terminology we introduce there. I will assume that you become sufficiently familiar with your LATEX distribution to be able to perform the editing cycle with the sample documents.
490
A.1
Appendix A
Installation
LATEX on a PC On a PC, most mathematicians use MiKTeX and the editor WinEdt. So it seems appropriate that we start there.
A.1.1
Installing MiKTeX If you made a donation to MiKTeX or if you have the TEX Live 2007 (or later) from the TEX Users Group, then you have a CD or DVD with the MiKTeX installer. Installation then is in one step and very fast. In case you do not have this CD or DVD, we show how to install from the Internet. To begin, go to the MiKTeX home page: http://miktex.org and under the Download/Install click on the version you want to install.1 You are directed to the MiKTeX download page, where you click on Download MikTeX Net Installer. This takes you to a list of the download sites, called “mirrors”. Choose one geographically close to you and click Download next to your pick. You are asked whether to Run or Save the installer application. Choose Save to save and now you have the setup application on your computer. Run setup and the MiKTeX Setup Wizard should start automatically. Then click Next and choose the task, Download only. Click Next again to choose the size of the download and choose Complete MikTeX. Again you have to choose a download site, and click Next a few more times, then Start, and the download starts. When it is complete, almost 35,000 files later, click Close. Now you have the files you need in the next step. The next task is installation. Run setup again, and up comes the Wizard. Click Next, and the task Install MikTeX is selected for you. Click Next, make sure you select Complete MikTeX. Click Next a few more times, select the default paper size, click Start, and the installation starts. When it is finished, click Close.
A.1.2
Installing WinEdt You can download WinEdt from its Web site and use it for 30 days before you pay the license fee. We now install WinEdt from the CD you are sent after you pay the license fee. Go to License and Registration at http://www.winedt.com Put the WinEdt installer CD in the DVD drive. The WinEdt Setup Wizard starts automatically. After accepting the licence, click Next a few times until WinEdt is installed and then click Finish. After installation, the WinEdt Configuration Wizard 1 We
follow the instructions for MiKTeX 2.5. Hopefully, this will also assist you with later versions.
A.1
LATEX on a PC
491
starts automatically. Click on the File Associations tab and click on Modify file type associations... under Current User, which is down the right side of the window, and then click OK. This gives all TEX files a lion icon and automatically associates them with WinEdt so that double clicking a TEX file automatically opens it in WinEdt. Clicking OK to close the Wizard. WinEdt’s claim to fame is its incredible customizability. Once you become familiar with the basic operations, you can make WinEdt behave the way you like.
A.1.3 The editing cycle In Section 1.2 you created the work folder for your work files. Start WinEdt by double clicking the WinEdt icon and open the file note1.tex in work, see the top half of Figure A.1. Observe: In the righthand corner, under the X button, the close application button, there is a small black x. All windows of WinEdt have such an x, this is the close window button. There are two rows of icons. The seventh from the right in the first row, a darkened lion’s head, is the TeXify button, use it to typeset your LATEX file. If you are not sure what an icon represents, let the cursor hover over it, and a brief description appears. A blue arrow in the left margin points at the line where the cursor is. Click on the TeXify icon. MiKTeX typesets note1.tex and produces another file, note1.dvi. The new file is displayed by the application Yap, which was automatically installed for you with MiKTeX. Rearrange the WinEdt and Yap windows. You should get an arrangement similar to the bottom half of Figure A.1.
A.1.4 Making a mistake WinEdt tells you if there is a mistake in your source file. To see what happens, click on the WinEdt window, and add a \ in line 11 of note1.tex, so that personally reads \personally. This makes \personally a command, which is a mistake (see Chapter 2). Click on the TeXify icon. We get the TeXify ... window, as in the top part of Figure A.2. At the ? prompt, type x for “exit” and press Return.2 You now see three windows, as shown in the bottom half of Figure A.2. The Yap window is mostly covered up. There is a new window, note1.log, the log window, in which the information from the TeXify ... window is recorded. In the WinEdt window, the blue arrow line pointer on the left indicates the offending line, and the mistake, \personally, is highlighted in red. Correct the mistake by deleting \, click on TeXify, and you are back in business. 2 The user, recorded on Figure A.2, typed s for “scroll mode”. LAT X then completes the typesetting E without stopping for errors (see Section D.4).
492
Appendix A
Figure A.1: note1.tex opened in WinEdt and typeset.
Installation
A.1
LATEX on a PC
Figure A.2: The mistake identified and localized (showing the log window).
493
494
A.1.5
Appendix A
Installation
Three productivity tools Now we see how the three productivity tools introduced in Section 1.4 are implemented in WinEdt and MiKTeX. Synchronization This is known as inverse search in WinEdt and MiKTeX. To set it up, open Yap, for example, by Texifying the document note1.tex. Choose the menu item View>Options>Inverse DVI Search. A list now displays the editors detected. Select WinEdt, click Apply, and click OK. You are done. To jump from the typeset file in Yap to the source file in WinEdt, double click a word in the typeset file. To jump from the source file to the typeset file, click on the DVI Search icon, the magnifying glass with the green leaf attached to the handle. You then jump to the typeset file, where a marker indicates the beginning of the typeset version of the source line the cursor is on. Block comment Select a number of lines in a source document and choose the menu option Edit>Move/Fill>Insert comment. Block uncomment is done with Edit>Move/Fill>Remove comment. Go to line This is done with Search>Go to Line... The icons and menu options all have keyboard equivalents. For instance, Go to line is Ctrl+G and TeXify is Ctrl+Shift and then press x.
A.1.6
An important folder When using MikTeX, where do you put new style files and packages? You can always place these files in your working folder. This is the simplest, but they will not be available when working with files outside that folder. If you want these files accessible everywhere (from all folders), then make a new folder in the LocalTeXFM folder tree, say C:\LocalTeXMF\MyStyles and place these files there. In order for MiKTeX to find these files, you now need to update the MiKTeX FileName Database. Do this: All Programs>MikTeX>MikTeX Options>Refresh FNDB or access it directly via the drop down menus from within WinEdt: Accessories>MikTeX Options The best way to install new packages from CTAN is via the Package Manager, which is accessed via
A.2
495
LATEX on a Mac All Programs>MikTeX>Package Manager
The Package Manager downloads the files from a CTAN mirror of your choice, then installs all files in their correct folders and updates MikTeX’s filename database. It is a good idea to run the MiKTeX Update Wizard on a regular basis to keep your system up to date: All Programs>MikTeX>MikTeX
A.2
Update Wizard
LATEX on a Mac
A.2.1 Installations
For the Mac, we install MacTeX,3 which consists of the TUG’s TEX Live and Richard Koch’s TeXShop. If you have TEX Live (see Section E.2), put the DVD in your computer’s DVD drive and follow the simple instructions. In a few minutes you are done. Otherwise, go to http://www.tug.org/mactex/ and in the Downloading section, click on MacTeX2007. After downloading about 700 MBs, you get the MacTeX2007 “disk image” that contains the mactex installer package. The disk image should open and the application MacTeX2007.mpkg should start automatically. A few more clicks—as in all Mac installations—and you are done. The spelling checker Finally, get the spell checker cocoAspell by Anton Leuski by going to http://people.ict.usc.edu/~leuski/cocoaspell/home.html and clicking on the download link, cocoAspell. This downloads the disk image cocoAspell.dmg, containing the installation package cocoAspell.mpkg. Follow the same process as above to mount this. Then double click on the installer package and follow the instructions. In the Apple menu, choose System Preferences..., where you find a new one, Spelling. Double click on it, and choose a dictionary. I use the dictionary English(United States). You should also select the filters you need. They are explained on the page you obtain by clicking on the Filter button. I selected Texinfo and TeX/LaTeX. You may have to restart the computer for the spelling checker to work. To invoke the spelling checker, select a word and press Command+Shift and :. It suggests a correct spelling. You can also add words to the dictionary. To learn more about the dictionaries, read the documents in /Library/Application Support/cocoAspell/ aspell6en6.00/doc/ 3 We
follow the instructions for MacTeX2007. Hopefully, this will also assist you with later versions.
496
A.2.2
Appendix A
Installation
Working with TeXShop Custom settings In due course, you can fully customize TeXShop as an advanced user. But to begin, there are just a few things to do. In the TeXShop menu choose Preferences... to open the TeXShop Preferences. To set the default font for the source files in TeXShop, click on the Document tab. Under Document Font click Set..., which brings up the Font window. Choose a font and size for the source files that is easy on your eyes. I use Courier and 12. Close the Font window. Make sure that under Editor all items except Auto Complete are selected. Now click the Preview tab and in the Preview Window Magnification enter a number for the magnification to be used for viewing the typeset version—I use 150. Once you type in the number, press Set. For Default Mouse Mode, choose Select Text. For Default Page Style, choose MultiPage. Under After Window Resize, choose Fixed Magnification. Then click on OK. Close the TeXShop Preferences window. When you become more familiar with TeXShop, you may want to revisit these settings. Changing a document for TeXShop If you use the graphicx package, see intrart.tex in the samples folder as an example, and your document contains illustrations that have been saved as EPS graphics and included with the \includegraphics command, then add the line \usepackage{epstopdf} in the preamble below the line \usepackage{graphicx} If you have many illustrations, it is preferable to open all the illustrations with TeXShop or Preview and save them in PDF format. Also, make sure that the extensions are not given in the \includegraphics commands, that is, \usepackage{graphicx} \includegraphics{products.eps} is changed to \usepackage{graphicx} \includegraphics{products}
A.2
LATEX on a Mac
Figure A.3: The document note1.tex: the source and the typeset version.
497
498
A.2.3
Appendix A
Installation
The editing cycle In your Document folder, you created the work folder (see page 4). We are going to work with the document note1.tex in the work folder. To start TeXShop, double click on TeXShop.app in the Applications folder, select the menu File>Open..., and navigate to the folder Documents/work/note1.tex Open the document. In the upper left corner of the source window, click the Typeset button. A second window opens, the preview window, showing note1.pdf, the typeset version of note1.tex. Unlike WinEdt, discussed in Section A.1, which produces a file called note1.dvi, TeXShop produces a PDF file, note1.pdf. Figure A.3 shows the two windows. At the top, you see TeXShop’s seven menus. For this introduction we ignore all but two menu options. You should use the Help menu to learn more and the Macros Help in the Help window along with the Macro Editor (open it with Macros>Open Macro Editor...) to become more productive.
A.2.4
Making a mistake TeXShop tells you if there is a mistake in your source file. Open note1.tex again and introduce a silly error, say, in the line \documentclass{amsart}, delete the closing brace, so it reads \documentclass{amsart. A new window—the third!—pops up, called note1 console, see Figure A.4. This is the log window (see Section 1.3). Click on the button Goto Error and the cursor is placed in the source document pretty close to the error. Now you can correct the error and typeset again.
A.2.5
Three productivity tools Now we see how the three productivity tools introduced in Section 1.4 are implemented in TeXShop. Synchronization Commandclick on a word in the source window. The preview window shows the corresponding typeset phrase circled in red. Similarly, commandclick on a word in the preview window and the corresponding source phrase is highlighted in yellow—it helps to click on text with no LATEX commands close by. Block comment Select a number of lines in a source document, and choose the menu option Format>Comment. All the lines, the whole block, are commented out. The reverse is done with Format>Uncomment. Go to line This is done with Edit>Line Number...
A.2
499
LATEX on a Mac
Figure A.4: The note1.tex console (log window). Buttons and menu options all have keyboard equivalents. For instance, Block comment is Command + { and Go to line is Command + L.
A.2.6 An important folder Create the texmf folder in the Library folder of your home folder—not the other Library folder, which is in the same folder as Applications. In texmf, create the tex folder, wherein you create the latex folder. Put all your personal (see Section 15.3) and additional sty files here.
APPENDIX
B Math symbol tables
B.1
Hebrew and Greek letters
Hebrew letters Type
Typeset
\aleph
ℵ
\beth
i
\daleth
k
\gimel
ג
502
Appendix B
Greek letters Lowercase Type
Typeset
Type
Typeset
Type
Typeset
\alpha
α
\iota
ι
\sigma
σ
\beta
β
\kappa
κ
\tau
τ
\gamma
γ
\lambda
λ
\upsilon
υ
\delta
δ
\mu
µ
\phi
φ
\epsilon
\nu
ν
\chi
χ
\zeta
ζ
\xi
ξ
\psi
ψ
\eta
η
\pi
π
\omega
ω
\theta
θ
\rho
ρ
\varepsilon
ε
\varpi
$
\varsigma
ς
\vartheta
ϑ
\varrho
%
\varphi
ϕ
\digamma z
\varkappa
κ
Uppercase Type
Typeset
Type
Typeset
Type
Typeset
\Gamma
Γ
\Xi
Ξ
\Phi
Φ
\Delta
∆
\Pi
Π
\Psi
Ψ
\Theta
Θ
\Sigma
Σ
\Omega
Ω
\Lambda
Λ
\Upsilon
Υ
\varGamma
Γ
\varXi
Ξ
\varPhi
Φ
\varDelta
∆
\varPi
Π
\varPsi
Ψ
\varTheta
Θ
\varSigma
Σ
\varOmega
Ω
\varLambda
Λ
\varUpsilon
Υ
B.2
503
Binary relations
B.2
Binary relations
Type < = \in \leq or \le \ll \prec \preceq \sim \simeq \equiv \subset \subseteq \sqsubseteq \smile \perp \mid \vdash \propto \bowtie \sqsubset \Join
Typeset < = ∈ ≤ ≺ ∼ ' ≡ ⊂ ⊆ v ^ ⊥  ` ∝ ./ < 1
Type
Typeset
> : \ni or \owns \geq or \ge \gg \succ \succeq \approx \cong \doteq \supset \supseteq \sqsupseteq \frown \models \parallel \dashv \asymp
> : 3 ≥ ≈ ∼ = . = ⊃ ⊇ w _ = k a
\sqsupset
=
Note the \colon command used in f : x → x2 , typed as f \colon x \to x^2
504
Appendix B
More binary relations Type
Typeset
Type
Typeset
\leqq
5
\geqq
=
\leqslant
6
\geqslant
>
\eqslantless
0
\eqslantgtr
1
\lesssim
.
\gtrsim
&
\lessapprox
/
\gtrapprox
'
\approxeq
u
\lessdot
l
\gtrdot
m
\lll
≪
\ggg
≫
\lessgtr
≶
\gtrless
≷
\lesseqgtr
Q
\gtreqless
R
\lesseqqgtr
S
\gtreqqless
T
\doteqdot
+
\eqcirc
P
\circeq
$
\triangleq
,
\risingdotseq
:
\fallingdotseq
;
\backsim
v
\thicksim
∼
\backsimeq
w
\thickapprox
≈
\preccurlyeq
4
\succcurlyeq
¶ † [ ] ♣ ♥ √
\hslash \backprime \Bbbk \diagup \blacktriangle \triangledown \square \lozenge \measuredangle \circledS \Finv
} 8 k N O ♦ ] s `
£ 2 0
Type
Typeset
\ell \jmath \partial \Re \prime \varnothing \exists \triangle \bot \S \ddag \natural \angle \diamondsuit \spadesuit \nabla \neg or \lnot \Diamond
` ∂ < 0 ∅ ∃ 4 ⊥ § ‡ \ ∠ ♦ ♠ ∇ ¬ 3
\complement \nexists
{ @
\diagdown \blacktriangledown \eth \blacksquare \blacklozenge \sphericalangle \bigstar \Game
H ð ^ F a
B.6
509
Delimiters
B.6
Delimiters
Name
Type
left parenthesis right parenthesis left bracket right bracket left brace right brace backslash forward slash left angle bracket right angle bracket vertical line double vertical line left floor right floor left ceiling right ceiling upward double upward downward double downward upanddown double upanddown upperleft corner upperright corner lowerleft corner lowerright corner
( ) [ or \lbrack ] or \rbrack \{ or \lbrace \} or \rbrace \backslash / \langle \rangle  or \vert \ or \Vert \lfloor \rfloor \lceil \rceil \uparrow \Uparrow \downarrow \Downarrow \updownarrow \Updownarrow \ulcorner \urcorner \llcorner \lrcorner
Typeset ( ) [ ] { } \ / h i  k b c d e ↑ ⇑ ↓ ⇓ l m p q x y
510
B.7
Appendix B
Operators
“Pure” operators, with no limits Type
Typeset
Type
Typeset
Type
Typeset
Type
Typeset
\arccos \arcsin \arctan \arg \cos \cosh
arccos arcsin arctan arg cos cosh
\cot \coth \csc \deg \dim \exp
cot coth csc deg dim exp
\hom \ker \lg \ln \log \sec
hom ker lg ln log sec
\sin \sinh \tan \tanh
sin sinh tan tanh
Operators with limits Type
Typeset
Type
Typeset
\det \gcd \inf \lim \liminf \injlim
det gcd inf lim lim inf inj lim
\limsup \max \min \Pr \sup \projlim
lim sup max min Pr sup proj lim
\varliminf
lim
\varlimsup
lim
\varinjlim
lim −→
\varprojlim
lim ←−
B.7
511
Operators
B.7.1 Large operators Type
Inline Rb
\int_{a}^{b}
a
a
\prod_{i=1}^{n} \coprod_{i=1}^{n} \bigcap_{i=1}^{n} \bigcup_{i=1}^{n} \bigwedge_{i=1}^{n} \bigvee_{i=1}^{n} \bigsqcup_{i=1}^{n} \biguplus_{i=1}^{n} \bigotimes_{i=1}^{n} \bigoplus_{i=1}^{n} \bigodot_{i=1}^{n} \sum_{i=1}^{n}
a
a
a
RRRR b a
R
···
Rb a
b
a
ZZZ
RRR b
\iiint_{a}^{b}
b
ZZ
RR b
\iint_{a}^{b}
\idotsint_{a}^{b}
a
I
Hb
\oint_{a}^{b}
\iiiiint_{a}^{b}
Displayed Z b
a
ZZZZ
Z ···
Qn `n
i=1 n a
Tn
i=1 n \
Sn
i=1 n [
Vn
i=1 n ^
Wn
i=1 n _
Fn
i=1 n G
Un
i=1 n ]
Nn
n O
Ln
n M
Jn
n K
Pn
n X
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
i=1
b
a
Z
n Y
i=1
b
i=1
i=1
i=1
i=1
i=1
a
b
512
B.8
Appendix B
Math accents and fonts
Math accents amsxtra Type \acute{a} \bar{a} \breve{a} \check{a} \dot{a} \ddot{a} \dddot{a} \ddddot{a} \grave{a} \hat{a} \widehat{a} \mathring{a} \tilde{a} \widetilde{a} \vec{a}
Typeset a ´ a ¯ a ˘ a ˇ a˙ a ¨ ... a .... a a ` a ˆ b a ˚ a a ˜ e a ~a
Type
Typeset
\spbreve \spcheck \spdot \spddot \spdddot
. .. ...
\sphat
b
\sptilde
∼
˘
∨
Math fonts Type LATEX \mathbf{A} \mathcal{A} \mathit{A} \mathnormal{A} \mathrm{A} \mathsf{A} \mathtt{A} \boldsymbol{\alpha} \mathbb{A} \mathfrak{A} \mathscr{a}
Typeset A A A A A A A α A A A
\mathscr requires the eucal package with the mathscr option
B.9
513
Math spacing commands
B.9
Math spacing commands
Name
Width
Short
Long
1 mu (math unit)
\mspace{1mu}
thinspace
\,
\thinspace
medspace
\:
\medspace
thickspace
\;
\thickspace
interword space
\
1 em
\quad
2 em
\qquad
Negative space 1 mu thinspace
\mspace{1mu} \!
\negthinspace
medspace
\negmedspace
thickspace
\negthickspace
APPENDIX
C Text symbol tables
C.1
Some European characters
Name
Type
Typeset
Type
Typeset
aring
\aa
˚ a
\AA
˚ A
aesc
\ae
æ
\AE
Æ
ethel
\oe
œ
\OE
Œ
eszett
\ss
ß
\SS
SS
inverted question mark
?‘
¿
inverted exclamation mark
!‘
¡
slashed L
\l
l
\L
L
slashed O
\o
ø
\O
Ø
516
C.2
Appendix C
Text accents
Name
Type
acute
\’{o}
breve
Typeset
Name
Type
o ´
macron
\={o}
¯o
\u{o}
o ˘
overdot
\.{g}
g˙
caron/haˇcek
\v{o}
o ˇ
ring
\r{u}
˚ u
cedilla
\c{c}
¸c
tie
\t{oo}
o o
circumflex
\^{o}
o ˆ
tilde
\~{n}
n ˜
dieresis/umlaut
\"{u}
u ¨
underdot
\d{m}
double acute
\H{o}
o ˝
underbar
\b{o}
grave
\‘{o}
o `
m . o ¯
dotless i
\i
ı
dotless j
\j
\’{\i}
´ı
\v{\j}
ˇ
C.3
Text font commands
C.3.1
Text font family commands
Command with Argument
Command Declaration
Typeset
Switches to the font family
\textnormal{...}
{\normalfont ...}
document
\emph{...}
{\em ...}
emphasis
\textrm{...}
{\rmfamily ...}
roman
\textsf{...}
{\sffamily ...}
sans serif
\texttt{...}
{\ttfamily ...}
typewriter style
\textup{...}
{\upshape ...}
upright shape
\textit{...}
{\itshape ...}
italic shape
\textsl{...}
{\slshape ...}
slanted shape
\textsc{...}
{\scshape ...}
small capitals
\textbf{...}
{\bfseries ...}
bold
\textmd{...}
{\mdseries ...}
normal weight and width
C.3
517
Text font commands
C.3.2 Text font size changes Command
LATEX sample text
AMS sample text
\Tiny
[not available]
sample text
\tiny
sample text
sample text
\SMALL or \scriptsize
sample text
sample text
\Small or \footnotesize
sample text
sample text
\small
sample text
sample text
\normalsize
sample text
sample text
sample text
sample text
\large \Large
sample text
sample text
\LARGE
sample text
sample text
\huge \Huge
sample text
sample text
sample text sample text
518
C.4
Appendix C
Additional text symbols
Name
Type
Typeset
ampersand
\&
&
asterisk bullet
\textasteriskcentered
∗
backslash
\textbackslash
\
bar (caesura)
\textbar

brace left
\{
{
brace right
\}
}
bullet
\textbullet
•
circled a
\textcircled{a}
a
circumflex
\textasciicircum
ˆ
copyright
\copyright
c
dagger
\dag
†
double dagger (diesis)
\ddag
‡
dollar
\$
$
double quotation left
\textquotedblleft or ‘‘
“
double quotation right
\textquotedblright or ’’
”
em dash
\textemdash or 
—
en dash
\textendash or 
–
exclamation down
\textexclamdown or !‘
¡
greater than
\textgreater
>
less than
\textless
math op.), 506 division, 22, 63, 158, 191, 506 \DJ (Dyet Eur. char.), 519 \dj (dyet Eur. char.), 519 doc (LATEX distr. directory), 311 document (text env.), 11, 37, 41, 68, 164, 248 document class options 9pt, 297 10pt, 104, 297, 306 11pt, 297, 306 12pt, 93, 297, 306 a4paper, 298, 306 a5paper, 306 are passed on to packages, 250 b5paper, 306 centertags, 298 combining, 308 draft, 13, 96, 249, 300, 306 executivepaper, 306 final, 300, 306 fleqn, 298, 307 for AMSFonts, 300 for bibliographies, 308, 308 for doublecolumn documents, 299, 307 for doublesided printing, 299, 307, 468 for font sizes, 297, 306 for page orientation, 307 for position of equation numbers, 298, 307 for start of chapters, 468 for title pages, 299, 307, 468 handout, 358 landscape, 307 legalpaper, 85, 298, 306 leqno, 298, 307 letterpaper, 298, 306 noamsfonts, 300 nomath, 300 notitlepage, 299, 307, 468 onecolumn, 299, 307 oneside, 299, 307, 468
574
DO–DO document class options (continued) openany, 468 openbib, 308, 308 openright, 468 psamsfonts, 300 reqno, 298, 307 tbtags, 298 titlepage, 249, 299, 307, 468 trans, 358 twocolumn, 85, 101, 259, 299, 307, 308, 313 twoside, 85, 249, 299, 307, 308, 468 document classes, 53, 82, 125, 376, 528, 529 amsart, 36, 82, 118, 125, 249–251, 254, 271–302, 465, 469, 525, 529, 536 amsbook, xxxii, 253, 465, 468, 469, 476, 525, 529 amsproc, 273, 529 anatomy of, 35–43 article, 35–43, 249, 251, 254, 263, 265, 273, 302, 303, 303–308, 525, 529 beamer, 325–359 book, xxxii, 253, 465, 469, 474, 524, 525, 529 changing, 536 cls files, 250, 375, 410 design of, 522 determine placement of equation numbers, 298, 307 position of equations, 298, 307 spacing, 102 foils, 53 for books, 253, 465, 466–473, 474, 476, 524, 525, 529 options of, 468 for presentations, see under FoilTEX and beamer legacy, 54, 118, 249, 254, 263, 273, 302, 305, 303–313, 529 letter, 308–310, 529 options of, see document class options proc, 529 procl, 272 report, 253, 303, 303–308, 529 sample, 10 sample.cls, 11 slides, 303, 529 document font families, 8, 14, 90, 89–90, 93, 123, 163, 301, 390, 412, 516, 544 normal, 90 roman, 89, 544 sans serif, 89, 544 typewriter style, 89, 544 document parameters, for page design, 552
Index documentation for AMS document classes and packages, 552 for beamer document class, 325, 358 for LATEX, 311, 552 packages, 312 for New Font Selection Scheme (NFSS), 553 \documentclass (preamble com.), 4, 13, 35, 36, 53, 70, 85, 180, 249, 294, 326, 355 and \NeedsTeXFormat command, 250 documents body of, 11, 36, 37, 41, 68, 164, 248–268 crossreferencing among multiple, 313 design of, 268–270 LATEX, 247–270 on the Internet, 317–323 using AMS document classes in, 271 legacy, 319 master, 477–479 preamble of, 249–251 printing and viewing, 5, 318, 530 readability of, 93, 358 splitting into multiple files, 476–479 dollar sign ($), see also $ and \$, 9 as inline math delimiter, 17, 18, 33, 74, 152, 152, 518 as text symbol, 77, 81 Doob, Michael, 110, 528, 557 \dot (x˙ math accent), 177, 512 . \doteq (= binary rel.), 503, 504 \dotfill (space fill com.), 106 dotless i and j (ı and ), 79, 79, 516 \dotplus (u math op.), 506 \dots (. . . math ellipsis com.), 24, 78, 160–161, 186 dots, filling lines with, 106 \dotsb (· · · math com.), 161 \dotsc (. . . math com.), 161 \dotsi (· · · math com.), 161 \dotsm (· · · math com.), 161 \dotso (. . . math com.), 161 double accents, in math, 177 double acute (˝ text accent), 79, 516 double dagger (‡) math symbol, 506, 508 text symbol, 81, 518 double guillemet, 519 double quote ("), 9, see also quotation marks, 75, 81, 518 in \index commands, 457 in B IBTEX database fields, 424, 445, 446 key, 9, 63, 75 double spacing, 99
Index
DO–EN double subscripts and superscripts, font size of, 199 double vertical lines in CD environments, 241 doublecolumn documents and footnotes, 313 document class options for, 85, 299, 307 figures and tables, 259 doublesided printing and marginal comments, 112 document class options for, 85, 299, 307, 468 \doublebarwedge ([ math op.), 506 \doublecap (e math op.), 506 \doublecup (d math op.), 506 \doublespacing (setspace com.), 99 \Downarrow (⇓) math arrow, 507, 509 math delimiter, 167 \downarrow (↓) math arrow, 507 math delimiter, 167, 509 \downdownarrows ( math arrow), 507 Downes, Michael, 447, 523, 524 \downharpoonleft ( math arrow), 507 \downharpoonright ( math arrow), 507 draft (doc. class opt.), 13, 96, 249, 300, 306 drivers printer, 312, 320, 526, 530, 539 video (DVI viewer), 526, 530 Duchier, Denys, 523 Dupr´e, Lyn, 85, 553, 557 DVI files, 526, 530, 539 converting to PostScript, 318 printing, 530 viewing, 526, 530 dvipdf (opt. of hyperref pack.), 320 dvipdfm (opt. of hyperref pack.), 320 dvipdfmx (opt. of hyperref pack.), 320 dvips (opt. of hyperref pack.), 320 dvipsnam (opt. of xcolor pack.), 349 dvipsone (opt. of hyperref pack.), 320 dviwindo (opt. of hyperref pack.), 320 dyet (Eur. char), 519 E email, see \email email clients, 65 editing cycle, 4–6 Mac, 498 PC, 491 edition (B IBTEX database field), 425 editor (B IBTEX database field), 425 editors, text, xxvi, 4, 5, 9, 18, 48, 55, 63, 84, 88, 365, 490, 526
575 Eijkhout, Victor, 553, 557 electronic mail address, see \email \ell (` math sym.), 508 ellipses (. . . ) centered (· · · ), 24, 160 diagonal, 235 in math, 24, 160–161, 233 in text, 78 vertical, 235 \em (font shape com. dec.), 90, 91, 92, 123, 367, 382, 516 em (font shape env.), 123 em (rel. unit), 403, 513 em dash (—), 14, 75, 81, 83, 518 EM fonts, see European Modern fonts \email (top matter com.), 278, 281–283, 295, 296 rules for using, 278 \emph (font shape com.), 8, 14, 68, 90, 91, 337, 516 emphasized (font shape), 8, 14, 28, 68, 90, 91, 92, 123, 125, 156, 367, 382, 516 and italic correction, 92 context dependence of, 91 empty (page style), 268 empty group ({ }), 39, 55, 70, 159, 179, 188, 189, 194, 204, 279, 309 \emptyset (∅ math sym.), 508 en dash (–), 14, 81, 83, 518 in B IBTEX databases, 427 Encapsulated PostScript (EPS), see EPS \enclname (redef. name), 376 encodings, font, see font encodings \end (end of env.), 67, 68, 72 errors with, 72, 146, 478 endofline characters, see line ending characters \endinput (input ending com.), 392, 478 eng (Eur. char), 519 \enlargethispage (spacing com.), 101, 484, 487 \enlargethispage* (spacing com.), 101 \ensuremath (math mode com.), 369, 370, 375 Enter key, see Return key entries bibliographic, see under bib glossary, see under glossaries index, see under indexing enumerate (list text env.), 313, 382 enumerate (pack.), 123, 313, 382, 387 enumi (counter), 123, 399 enumii (counter), 399 enumiii (counter), 399 enumiv (counter), 399
576
EN–ER environments, 8, 14, 18, 67–74 *ed forms, 157, 201, 211, 213, 215, 231, 259 act as braces, 72 arguments of, see under arguments or specific environments \\ in, 124, 145 begin with \begin, 68, 72 body of, 68 casesensitivity of names of, 69 end with \end, 68, 72 font size, 123 for presentations, see under beamer and FoilTEX for proclamations, see proclamations for tables, see tables indenting contents of, in source file, 155 legacy, 123, 143–145, 407 list, see list text environments logical design using, 535 typesetting environment names, 535 math, see under inline and displayed math environments, subsidiary math environments, and displayed text environments modifying, 380–383 \newline in, 98 operating, see under PC, Mac, and UNIX short, 385 subsidiary math, see subsidiary math environments text, see displayed text environments, list text environments, and text environments userdefined, see userdefined environments, see userdefined environments visual design using typesetting environment names, 535 EPS (Encapsulated PostScript) files, 476, 483, 496 format, 39, 259, 483, 552 \epsilon ( Greek char.), 502 \eqcirc (P binary rel.), 504 eqnarray (math align. env.), 222 \eqref (crossref. com.), 28, 29, 31, 156, 202, 203, 215, 219, 255, 256, 263, 328, 370 \eqslantgtr (1 binary rel.), 504 \eqslantless (0 binary rel.), 504 equals (=) binary relation, 63, 503 in B IBTEX database fields, 424 key, 8 equation (counter), 399, 399
Index equation (disp. math env.), 27, 33, 156–157, 201, 231, 328 blank lines in, 157 equation* (disp. math env.), 157, 157, 201, 231, 328 \equationname (hyperref redef. name), 323 equations, 27–29 document class options for placement of numbers, 298, 307 grouping, 203–204 in multiline math environments, 215 in presentations, 328 labels for, 28, 256 marginal comments in, 112 numbering of, 27–29, 55, 156–157, 201, 211, 213, 215 groups, 218–219 in chapters of books, 466 preventing, 30, 31, 157, 201, 211, 213, 215, 218, 231 variants, 203, 218 within sections, 156, 250 position of, 298, 307 systems of, 225–226 tagging, 29, 201–204, 211, 213, 215, 218, 230, 231 \equiv (≡) binary relation, 503 math operator, 24, 174, 182, 371, 379 error messages, 5, see also warning messages, 48–52, 534 Argument of \xxx has an extra }, 146 Bad math environment delimiter, 132, 153, 475 \begin{document} ended by \end{xxx }, 87, 146 \begin{split} not allowed here, 231 \begin{xxx } on input line xx ended by \end{yyy }, 49, 50, 72 Can be used only in preamble, 250 Command \xxx already defined, 367, 376, 381 Display math should end with $$, 51 Double superscript, 158 Environment xxx undefined, 381 Extra alignment tag has been changed to \cr, 237 Extra }, or forgotten \xxx , 73, 223 File ‘xxx ’ not found, 534 I was expecting a ‘,’ or a ’)’, 445, 446
Index
ER–FI Illegal character in array arg, 237 Illegal unit of measure (pt inserted), 114 Incomplete \iffalse; all text was ignored after line xx , 309 Invalid use of \xxx , 162 line numbers in, 6, 9, 153, 154 Misplaced \xxx, 135, 227, 233 Missing \begin{document}, 126 Missing control sequence inserted, 404 Missing $ inserted, 20, 51, 147, 153, 154, 157, 233 Missing number, treated as zero, 98, 111 Missing \right. inserted, 217 Missing } inserted, 21, 49, 223 No counter ‘xxx ’ defined, 400 Paragraph ended before \xxx was complete, 49, 50, 114, 126, 237, 373, 379, 462 recorded in log file, 526 Runaway argument?, 49, 126, 237, 284, 379, 462 showing paragraph breaks in, 99 Something’s wrongperhaps a missing \item, 120, 263 \tag won’t work here, 215 Text line contains an invalid character, 9, 52, 64 Too many }’s, 21, 72, 285 Undefined control sequence, 44, 50, 71, 282, 284, 367 Underfull \hbox, 97, 530 Use of \xxx doesn’t match its definition, 379 \verb command ended by end of line, 148 with B IBTEX, 443–446 in log (blg) files, 442 You’re missing a field name, 444 eszett (ß, SS), 81, 515, 548 \eta (η Greek char.), 502 \eth (ð math sym.), 508 eth (Eur. char), 519 eucal (pack.), 301, 302, 386, 512 options, 301, 390, 512 eufrak (pack.), 302 Euler Fraktur (math alphabet), 180, 197, 301, 302, 390, 512 Euler Script (math alphabet), 301, 302, 390, 512 European accents, 15, 79–80
577 characters, 15, 79–80, 515–516, 519 quotation marks, 519 European Modern (EM) fonts, 524 ex (rel. unit), 403 examples of bibliographies, 40–41, 43, 47, 261–263, 421–423 of indexes, 450–452 of nonbreakable spaces (~ spacing com.), 76 of top matter commands, 282–285 \except (table of contents com.), 473 exclamation marks (!), 9, 62 as float control, 260, 486 in \index commands, 453, 454, 457 Spanish (¡), 81, 515, 518 terminating sentences with, 64–67 executivepaper (doc. class opt.), 306 exercises, 476 in books, 476 within a section, 476 \exists ( ∃ math sym.), 508 exiting, from interactive mode, 491, 534 \exp (exp math op.), 172, 510 expanding values of counters, 80, 400 exscale (pack.), 301, 311 extensibility of LATEX and TEX, 525 F Fairbairns, Robin, 523 \fallingdotseq (; binary rel.), 504 fancyhdr (pack.), 270, 537, 540 FAQ (Frequently Asked Questions), 539, 542 \fboxrule (length com.), 110 \fboxsep (length com.), 110 \fcolorbox (beamer com.), 349 fd (font def. file), 544, 545, 546 Fear, Simon, 140 fields, bibliographic, see under bib figure (counter), 399 figure (float env.), 39, 259–260, 327 optional arguments of, 260, 486 placement of, 484, 486 figure* (float env.), 259 \figurename (redef. name), 375, 376 and hyperref package, 323 figures, 259–260 captions in, 476 fragile commands in, 74 commands for, 259, 476, 479 doublecolumn, 259 forcing typesetting of, 101 lists of, see also under lists, 475–476 adding a line to, 476 fragile commands in, 74
578
FI–FO figures (continued) numbering of, 258, 259 oldstyle, 62 placement of, 260, 484, 486 file formats, for graphics EPS , 39, 259, 483, 496, 552 PDF , 39, 259, 326 file transfer protocol, see FTP fileerr (pack.), 313 files MakeIndex log, see ilg auxiliary, see aux B IBTEX databases, see bib B IBTEX log, see blg B IBTEX style, see bst class (cls), see under document classes command, see command files converting from articles to presentations, 53, 326 documents composed of multiple, 476–479 DVI , see DVI files Encapsulated PostScript, see EPS font definition, see fd font metric, see also tfm, see under font metrics glossary, see glo hyperref bookmarks, see out hyphenation pattern, see under hyphenation including in other files, 386, 478 index entry, see idx list of figures, see lof of tables, see lot listing those used by a document, 387 log, see log files naming of, 44 with \graphicspath command, 479 organization of, 476–479 package source, see under packages portability of, 9 Portable Document Format, see PDF, see PDF
PostScript (PS), see under PostScript processed bibliography, see bbl processed index, see ind sample, see sample files source, see source files start on new page with \include command, 478 style, see sty tables of contents, see toc terminating, 392, 478 filling horizontal space, 106, 113, 126, 186, 233, 235, 414
Index final (doc. class opt.), 300, 306 final preparation, of books, 482–487 fine tuning of mathematical formulas, 190–192 placement of root with \sqrt, 162 placement of text boxes, 115 \Finv (` math sym.), 508 fixedsize math delimiters, 168 fixltx2e.dtx, 311 fixltx2e.ins, 311 fixltx2e.sty, 311, 312 flalign (math align. env.), 208, 210, 221 \flat ( [ math sym.), 508 fleqn (doc. class opt.), 298, 307 floats, see also tables and figures, 258–261 commands, 259, 260 controls, 134, 260–261 environments, 133, 258–260 figure, 39, 259–260, 260, 327 figure*, 259 locating, 484, 486 table, 134, 258, 260 table*, 259 forcing typesetting of, 101 specifying placement of, 260, 484 floor math delimiters, 167, 509 flush left alignment of formulas, 209, 210, 212, 214, 221, 307 column alignment, 231–232, 236 columns in tabular environments, 135 setting equations, 298, 307 flush right alignment of formulas, 209, 210, 212, 214, 221 alignment of text, 106 column alignment, 231–232, 236 columns in tabular environments, 135 setting equations, 298, 307 flushleft (text align. env.), 8, 124, 407 breaking lines in, 124 flushright (text align. env.), 8, 68, 124, 407 breaking lines in, 124 \fnsymbol (footnote counter style com.), 87 \foilhead (FoilTEX com.), 53, 55 optional argument of, 56 foils (doc. class), 53 FoilTEX, 53 commands for, 55 folders, 3 samples, see samples structure of, 477–479 work, see work font command declarations, see also font commands, 90
Index
FO–FO for shape \em, 90, 91, 92, 123, 516 \itshape, 90, 91, 92, 94, 123, 412, 516 \normalfont, 90, 90, 123, 163, 516, 545 \rmfamily, 90, 123, 516 \scshape, 90, 91, 123, 516 \sffamily, 90, 94, 123, 516 \slshape, 90, 91, 92, 94, 95, 123, 516 \ttfamily, 90, 123, 142, 516 \upshape, 90, 91, 123, 382, 412, 516 for weight \bfseries, 71, 73, 90, 93, 94, 95, 123, 516 \mdseries, 90, 93, 516 font commands, see also font command declarations and italic correction, 90, 92 and MakeIndex, 90 for selecting fonts using family names, 544 for series \textmd, 90, 93, 516 for shape \emph, 14, 68, 90, 91, 337, 516 \textit, 90, 91, 337, 516 \textnormal, 90, 90, 163, 516 \textrm, 90, 301, 516 \textsc, 90, 91, 516 \textsf, 90, 516 \textsl, 90, 91, 516 \texttt, 14, 76, 90, 516 \textup, 90, 91, 516 \upn, 156 for size \footnotesize, 93, 94, 517 \Huge, 93, 93, 517 \huge, 93, 93, 517 \LARGE, 93, 93, 517 \Large, 93, 93, 94, 517 \large, 93, 93, 517 \larger, 94 \normalsize, 93, 93, 517 \scriptsize, 93, 93, 517 \SMALL, 93, 93, 517 \Small, 93, 93, 517 \small, 93, 93, 517 \Tiny, 93, 93, 517 \tiny, 93, 93, 517 for weight \textbf, 14, 73, 90, 93, 163, 301, 337, 516 lowlevel, 95 math, see math font commands
579 obsolete, 94 orthogonality of, 94 twoletter (obs. LATEX 2.09), and italic correction, 95 using in math environments, 163 font encodings, 78, 78 koi8u, 548 LY1, 545, 546 OT1, 544 T1, 82, 519, 547 font environments for series bfseries, 123 for shape em, 123 itshape, 123 rmfamily, 123 scshape, 123 sffamily, 123 slshape, 123 ttfamily, 123 upshape, 123 for weight bfseries, 123 font families, document, see document font families font metrics files, 318, 529, 544, 545 font series, 89, 93 command declarations for, see command declarations commands for, see under font commands environments for, see under font environments font shapes command declarations for, see under font command declarations commands for, see under font commands environments for, see under font environments sans serif, 89 serif, 89 font substitution, 94, 318, 484 warning messages, 94 font weight, 89 command declarations for, see under font command declarations commands for, see under font commands environments for, see under font environments font width, 89 fontenc (pack.), 545, 547, 548 \fontfamily (font selection com.), 544
580
FO–FR fonts AMSFonts, 300, 301–302, 525, 528 document class options for, 300 obtaining, 540 PostScript versions of, 300, 301, 524, 525, 533 technical support for, 542 updates to, 533 bitmap, 543 blackboard bold, 181, 197, 390 bold math, 184, 196, 196, 197, 512 calligraphic, 185, 197, 301, 390, 512 commands, see under font commands and font command declarations Computer Modern, see Computer Modern Courier (typewriter shape), 317, 496, 544 encodings, see font encodings Euler Fraktur, 180, 197, 301, 302, 390, 512 Euler Script, 301, 302, 390, 512 European Modern (EM), see European Modern fonts for text in math mode, 163 Helvetica, 89, 544 including in PDF files, 318 including in PostScript files, 317 Lucida Bright, 543, 546 math, 195–197 size of, 199 monospaced, 88 names of LATEX 2.09, 313 PSNFSS (Berry scheme), 544 PostScript, see under PostScript proportional, 88 proprietary using in PDF files, 318 using in PostScript files, 317 samples, 313 selecting, 88–95 with \fontfamily command, 544 with \selectfont command, 95, 544 shapes, see font shapes size of, 55, 89, 93, 93–94, 123, 517 document class options for, 297, 306 sources for, 546 substitution of, 94, 318, 484 Times, 89, 544–545 typewriter style, 88 weight, see font weight width, see font width fontsmpl (pack.), 313 fonttbl.tex (sample file), 78, 134 footers, page, 268–270 \footnote, 87 footnote (counter), 399
Index footnotes, 87–88 and doublecolumn documents, 313 fragile commands in, 74 in minipage environments, 112 indicated by symbols, 87 on title pages, 87 unmarked, 44, 279 \footnotesize, 93, 94, 517 \forall (∀ math sym.), 508 forcing floats to typeset, 101 indentation, 99 formats LATEX, 529 TEX creating with initex, 528 files read by virtex, 528 Plain TEX, 528 formats, file, see file formats formatting of documents, determined by document classes, 61 formula.tex (sample file), 164–166 formulas, 18 adjusted, 210, 231–242 flush left and right, 214 aligning, 30–32, 209–210, 219–231, 242, 298, 524 multiline, 217–218 rules for, 217–218, 221 text in, 226–227 annotating, 225 boxed, see under boxes breaking into multiple lines, 30–31, 217–218 displayed, 483 rules for, 217 commas in, 155 components of, 22–27, 157–162 building using, 164–166 displayed, see displayed math environments gallery, 180–186 grouping, 210, 211–212 inline, see inline math environments multiline, 30–32 splitting, 212–214 numbering, see under equation specifying fonts for, 163 foundries, type, 544, 546 ), 19, 21, 23, 70, 158, 186, 200 \frac ( x y fractions, 19, 21, 23, 70, 158, 186 continued, 200 displayed, 158, 186 generalized, 204–205 inline, 158
Index
581
FR–GT fragile commands, 74 protecting, 74, 252, 253, 466, 475 frame (beamer env.), 327, 328, 344 \framebox, 109 frames, see also under presentations, 327, 358 environments of, 327 outline (table of contents), 328 titles, 327, 358 \frametitle (beamer com.), 327, 328, 344 \frenchspacing (spacing com.), 67 Frequently Asked Questions (FAQ), 539, 542 front matter, 467 numbering of chapters in, 467 of AMS documents, 285 of books, 473–475 \frontmatter (struct. com.), 467 \frown (_ binary rel.), 503 ftnright (pack.), 313 FTP (File Transfer Protocol), 538 G \Game (a math sym.), 508 \Gamma (Γ Greek char.), 502 \gamma (γ Greek char.), 502 gather (disp. math env.), 208, 210, 211–212, 219, 483 page breaks in, 242 rules for, 211 gather* (disp. math env.), 211 gathered (subsid. math env.), 211, 227–230 and \allowdisplaybreaks commands, 243 \gcd (gcd math op.), 172, 510 \ge (≥ binary rel.), 503 generalized commands, see userdefined commands fractions, 204–205 \genfrac (math com.), 204–205 geometry (pack.), 270 \geq (≥ binary rel.), 503 \geqq (= binary rel.), 504 \geqslant (> binary rel.), 504 german (opt. of babel pack.), 548 german.tex (sample file), 548 \gg ( binary rel.), 503 gg.tex (sample file), 44–48 gg2.tex (sample file), 45–46 ggamsart.tpl (template file), 297 \ggg (≫ binary rel.), 504 \gimel ( גHebrew char.), 501 glo (glossary files), 464, 531 global commands, see under scope glossaries, 464 as a custom list environment, 414 auxiliary file (glo), 464, 531
defining, 464, 531 \glossary (glossary com.), 464, 531 glue, 378, 407, 530, 530 horizontal, 530 parameters of, 530 vertical, see also under vertical spacing, 530 glyphs, see also characters measuring, 530 \gnapprox ( neg. binary rel.), 505 \gneq ( neg. binary rel.), 505 \gneqq ( neg. binary rel.), 505 \gnsim ( neg. binary rel.), 505 go to line productivity tool, see jump to a line Google, 464, 538 Goossens, Michel, xxxiii, 558, 559 graphics commands for, 54, 259, 259, 260, 327, 341, 479, 496 file extensions in, 496 in presentations, 342 formats EPS , 39, 259, 483, 496, 552 PDF , 39, 259, 326 including in documents, 259–260 scaling, 259 graphics (LATEX distr. directory), 311, 539 graphics (pack.), 312 \graphicspath (graphics com.), 477 graphicx (pack.), 37, 39, 250, 259–260, 300, 523, 552 grave ( ` grave text accent), 62, 79, 516 \grave (` x math accent), 177, 512 greater than (>) as binary relation, 503 text symbol, 81, 518 Greek letters, 181, 389–390, 502 Greenwade, George D., 99 grouping chapters into parts, 466 equations, 203–204 formulas, 210, 211–212 symbols with math delimiters, 169 groups of equations, 203–204 crossreferencing, 218–219 labels for, 204 numbering, 218–219 of tokens, 530 \gtrapprox (' binary rel.), 504 \gtrdot (m binary rel.), 504 \gtreqless (R binary rel.), 504 \gtreqqless ( T binary rel.), 504 \gtrless (≷ binary rel.), 504
582
GT–HY \gtrsim (& binary rel.), 504 guillemets, 519 \guillemotleft (Eur. quot. mark), 519 \guillemotright (Eur. quot. mark), 519 \guilsingleft (Eur. quot. mark), 519 \guilsingright (Eur. quot. mark), 519 Gurari, Eitan, 558 GUTenberg (French TEX user group), 541 \gvertneqq ( neg. binary rel.), 505 H \H double acute text accent (˝), 79, 516 Hahn, Harley, 554, 558 handout (beamer doc. class opt.), 358 \hangafter TEX command, 377 length command, 100 \hangindent (length com.), 99 hanging indentation, 99–100, 145 Hargreaves, Kathryn A., 553, 557 Harrison, Michael A., 449, 463, 557 Hart, Horace, 553, 559 \hat (ˆ x math accent), 25, 177, 512 \hbar (~ math sym.), 508 \hdotsfor (space fill com.), 186, 233, 235 optional argument of, 233 headers, page, see running heads headings (page style), 268 \headtoname (redef. name), 376 \heartsuit (♥ math sym.), 508 Hebrew letters, 501 \height (length com.), 108, 108, 111 height, of text boxes, 108, 110, 111, 113, 405, 405 Helvetica (sans serif font), 89, 544 \hfill (space fill com.), 106, 126, 131, 414 \Hfootnotename (hyperref redef. name), 323 \hfuzz (\hbox warning adjustment), 96 hhline (pack.), 140, 313 hierarchy of structural commands, 252, 466 history of AMSLATEX, AMSTEX, TEX, and LATEX, 521–525 \hline (table com.), 135, 137 Høgholm, Morten, 206 \hom (hom math op.), 172, 510 \hookleftarrow (← math arrow), 507 \hookrightarrow (,→ math arrow), 507 horizontal lines (rules), in tabular environments, 135, 137 horizontal spacing, 313 adjusting, 188 commands \ (interword space), 11, 66, 70, 102, 190, 520
Index \! (negthinspace), 182, 190, 190, 513, 520 \, (thinspace), 25, 67, 75, 188, 190, 190, 513, 520 \: (medspace), 190, 513, 520 \; (thickspace), 190, 513, 520 \@. (intersentence space), 520 \dotfill (fill com.), 106 \frenchspacing, 67 \hdotsfor (fill com.), 186, 233, 235 \hfill (fill com.), 106, 126, 131, 414 \hrulefill (fill com.), 106, 113 \hspace, 102, 142 \hspace*, 69, 103–104 \medspace, 190, 513, 520 \mspace, 190, 190, 513 \negmedspace, 190, 513, 520 \negthickspace, 190, 513, 520 \negthinspace, 182, 190, 190, 513, 520 \nobreakspace, 76 \nonfrenchspacing, 67 \phantom, 102–103, 191–192, 223, 406 \qquad, 27, 102, 105, 190, 190, 224, 513, 520 \quad, 24, 27, 102, 105, 154, 190, 190, 225, 513, 520 \thickspace, 190, 513, 520 \thinspace, 25, 67, 75, 188, 190, 190, 513, 520 to avoid using, 481 \xspace, 368–369 in math, 19, 154–156, 187–192, 233, 236 in text, 11, 64–67, 102–104, 114, 126, 414, 513, 520 intersentence, 11, 64–67 interword space ( ), 11, 64–67, 70, 76, 102, 383, 427 preventing removal of, 103 howpublished (B IBTEX database field), 425 \href (hyperref com.), 323 \hrulefill (space fill com.), 106, 113 \hslash (~ math sym.), 508 \hspace (spacing com.), 102, 142 \hspace* (spacing com.), 69, 103–104 HTTP (Hypertext Transfer Protocol), 538 \Huge (font size com.), 93, 93, 517 \huge (font size com.), 93, 93, 517 \hyperlink (beamer com.), 344 hyperlinks, 319 in PDF files, 319–323 increasing size of, 322 suppressing, 322 to bibliographic citations, 321 to Web sites, 322–323
Index
583
HY–IN in presentations, 343–347, 358 putting bookmarks in documents, 321 to bibliographic citations, 321 with hyperref pack., 320, 321, 320–322 hyperref (pack.), 320–323 bookmarks file (out), 321 commands \autoref, 320, 320, 322, 323 \pageref, 322 \ref, 322 \urladdr, 323 \WriteBookmarks, 322 documentation for, 320 options, 320 bookmarks=true, 320, 321 colorlinks, 320, 321, 321 dvipdf, 320 dvipdfm, 320 dvipdfmx, 320 dvips, 320 dvipsone, 320 dviwindo, 320 hypertex, 320 pagebackref, 320, 321, 321 pdftex, 320 ps2pdf, 320 tex4ht, 320 textures, 320 vtex, 320 hypertex (opt. of hyperref pack.), 320 Hypertext Transfer Protocol, see HTTP hyphen.tex, 82 hyphenation, 13, 82–85, 530 determined by optional hyphen, 82 displaying, 84 LATEX’s algorithm, 79, 82 of German text, 13, 548 of hyphenated words, 13 of words with accents, 82 preventing, 83–84, 107 rules for English, 85 specifying, 82 tables, 528 \hyphenation, 82 hyphens, 14, 62, 75 key, 9, 63 optional, 13, 82, 82, 96, 483 unbreakable, 83 I \i (ı dotless i), R 79,R79, 516 \idotsint ( ··· large math op.), 161, 175, 511 idx (index aux. file), 459, 460, 461, 464
\iff ( ⇐⇒ math arrow), 365, 507 \ignorespacesafterend (spacing com.), 383 RRRR \iiiint ( large math op.), 161, 175, 511
RRR RR large math op.), 161, 175, 511
\iiint (
\iint ( large math op.), 161, 175, 511 ilg (MakeIndex log files), 461 illustrations, 258, 259–260 commands for, 479, 496 doublecolumn, 259 Encapsulated PostScript (EPS), 39, 259, 476, 483, 496 graphicx (pack.), 250, 259 placement of, 102, 260 portability of, 552 Portable Document Format (PDF), 39, 259, 476, 496, 552 preparation of, 483 sample files, 39, 54, 327 scaling, 54, 259 with graphicx (pack.), 37 with the picture environment, 259 \Im (= math sym.), 508 \imath (ı math sym.), 508 implementations, LATEX, see LATEX implementations \in (∈ binary rel.), 503 in (inch abs. unit), 12, 15, 96, 105, 107, 111, 403 inbibl.tpl (sample file), 263 INBOOK (bibl. entry type), 424, 436 inch (in abs. unit), 12, 15, 96, 105, 107, 111, 403 \include (inclusion com.), 392, 477, 478 and auxiliary files, 531 and counters, 400 errors with, 478 files start on new pages, 478 \includegraphics (graphics com.), 54, 260, 327, 479, 496 file extensions in, 259, 496 in presentations, 341, 342 \includeonly (preamble inclusion com.), 400, 477, 478 including files in other files, 386, 392 fonts in PDF files, 318 in PostScript files, 317 graphics in documents, 259–260 INCOLLECTION (bibl. entry type), 424, 431 incrementing counters, 399 ind (proc. index files), 459, 460, 461, 487 \indent (indentation com.), 99
584
IN–IN indentation hanging, 99–100, 145 in multline and multline* environments, 213 in source files, 166 of environment bodies, 155 of first lines of paragraphs, 99, 313, 404 of subformulas, 217 of text, 99–100 customizing, 103 forcing, 99 preventing, 99 indentfirst (pack.), 313 \index (indexing com.), 371, 450, 450, 453–459 at signs (@) in, 455, 456 defining crossreferences in indexes with, 457 double quote (") in, 457 encapsulators, 454, 455 exclamation marks (!) in, 453, 454, 457 modifiers, 454, 455 page ranges in, 455 placement of commands, 458, 486 spaces in argument of, 462 syntax of, 458 vertical line () in, 454, 457 indexes, 311, 452, 449–464 appear in back matter, 450 auxiliary files (idx), 459, 461 crossreferences in, 457 defining, see indexing entries sorting, 455–456 stored in idx files, 459, 461 log files (ilg), 461 main entries, 267, 449, 453 multiple, 463–464 placement of, 450, 461 processed files (ind), 459–461, 487 processing, 459–461 sample, 452, 450–452 simple, 267 special characters in, 457 subentries, 267, 449, 453 subsubentries, 267, 449, 454 syntax of, 458 typesetting, 459–461 using showidx, 453 vertical space in, 267, 449 indexing, see also MakeIndex, 449–464 commands for, 267, 371, 449–459, 461 CTAN systems for, 539 entries main, 267, 449, 453
Index subentries, 267, 449, 453 subsubentries, 267, 449, 454 environment for, 267, 407, 449, 459 page ranges, 454–455 placement of commands for, 458 processed files (ind), 487 processed index files (ind), 459–461 rules for, 462–463 typesetting entries in margins, 311 with AMS document classes, 450 with userdefined commands, 371 \indexname (redef. name), 376 \indexspace (indexing com.), 449 \inf (inf math op.), 172, 299, 510 information, top matter AMS specific, 279–281 AMS subject classifications, 279 keywords, 280, 297 author, 275–279, 304 addresses, 44, 277, 296 current addresses, 277, 281 email addresses, 278, 281, 296 Internet addresses, 278, 281, 296 research support, 44, 279, 281, 296, 304 for articles, 37–38, 273–285 author names, 276, 296, 304 dedications, 275 title, 273, 297, 304 translator, 274 \infty (∞ math sym.), 20, 508 initex (TEX program), 528 initials periods in, 67 typographical rules for, 67 \injlim (inj lim math op.), 172 inline fractions, 158 math formulas, see inline math environments tables, 133 inline math environments, 17, 152 \( (start math mode), 18, 74, 152 \) (end math mode), 18, 74, 152 act as special braces, 152, 153 and $, 17, 18, 33, 74, 152, 152 displaystyle binomials in, 158 font size of, 199 math, 152 matrices in, 235 size of delimiters in, 170 INPROCEEDINGS (bibl. entry type), 424, 430–431 \input (inclusion com.), 528 inputenc (pack.), 547 \institute (beamer com.), 326
Index
585
IN–KE institution (B IBTEX database field), 425 instructions to LATEX, see also commands and environments, 7, 11, 35, 48, 63, 67, R 248, 252, 534 \int ( large math op.), 25, 175, 511 integrals, 25, 161, 299 integration of fonts into LATEX, 523, 543 interactive mode, 113, 376–378, 491, 534–535 \intercal ( math op.), 506 intercolumn spacing, 209, 220, 220 in aligned math environments, 224, 228 in tabular environments, 136 interline spacing, 242 adjusting, 95, 98–99, 104 international TEX users groups, 537 Internet, 278, 322, 415, 538 addresses, 279, 323 in top matter (\urladdr com.), 278, 279, 296, 323 and LATEX, 537–542 browsers, 538, 540 FTP transfers, 538 viewing PDF files in, 318 documents in PostScript format, 317 finding LATEX and TEX related material on, 537 finding LATEX and TEXrelated material on, 538 hyperlinks in PDF documents, 319, 323 LATEX and TEX resources on, 538, 542 putting LATEX documents on, 317–323 types of sites for downloading files, 538 intersentence spaces (\@.), 11, 64–67, 520 and \frenchspacing, 67 and \nonfrenchspacing, 67 \intertext (text in math com.), 226–227 interword space, 11, 64–67, 76, 383, 427 and \frenchspacing, 67 and \nonfrenchspacing, 67 command (\ ), 11, 66, 70, 102, 190, 520 intlimits (opt. of amsmath pack.), 299 intrart.tex (sample file), 35–43, 53, 55, 326, 331 typeset, 41–43 intrarti.idx (index entry file), 459, 460, 461 intrarti.ilg (index log file), 461 intrarti.ind (index proc. file), 459–461 intrarti.tex (sample file), 450–452, 460, 459–461 intropres.tex (sample file), 53, 55, 57–58 invalid characters, 9, 63 inverse search, see synchronization invisible boxes, see struts invoking delimited commands, 378
proclamations, 37, 46, 52, 125, 128, 343, 384 userdefined commands, 366 \iota (ι Greek char.), 502 ISBN (bibl. com.), 425 \it (obs. LATEX 2.09 font com.), 94 italic correction, 91–93, 520 and commas, 92 and \em, 92 and font commands, 90, 92 and \itshape, 92 and periods, 92 and \slshape, 92 suppressing, 92 with twoletter font commands, 95 italics bold math, 197 font shape, 14, 28, 68, 88, 90, 91, 94, 123, 412, 516 in math mode, 196, 196, 512 math font, 196 \item (list item com.), 118–123 and square brackets, 122 in simple indexes, 267, 449 optional arguments of, 120, 122, 409, 410 \itemindent (length com.), 407, 408 itemize (text list env.), 118, 119 \Itemname (hyperref redef. name), 323 \itemsep (length com.), 407, 413 \itshape (font shape com. dec.), 90, 91, 94, 123, 412, 516 and italic correction, 92 itshape (font shape env.), 123 J \j ( dotless j), 79, 79, 516 Jackowski, Bogusław, 525 Java, with B IBTEX, 447 JavaScript, 318 Jeffrey, Alan, 523 Jensen, Frank, 403 \jmath ( math sym.), 508 \Join (1 binary rel.), 503 Jones, David M., 447, 524 journal (B IBTEX database field), 425 Jr., in bibliographic entries, 426 jump to a line productivity tool, 6, 494, 498 justification of text left, 8 right, 8, 14, 68 K \k (ogonek Eur. accent), 519 \kappa (κ Greek char.), 502 \ker (ker math op.), 172, 510
586
KE–LA Kern, Uwe, 325, 348, 349, 559 kerning, 529, 530 key (B IBTEX database field), 425 keyboard, 8–9, 11, 62–64, 547 equivalents, 494, 499 keys, 17 alphanumeric, 8–9 arrow, 53 Enter, see Return key Esc (escape), 63, 365 letter, 8–9, 62 math symbol, 63 math typing, 17 modifier, 62, 63 number, 62 prohibited, 9, 63 Return, 9, 11, 63 space bar, 9, 11, 53, 63, 325, 334 special, 9, 17, 63, 76 Tab, 9, 11, 63 keys, sorting (in \index commands), 455–456 case sensitivity of, 462 \keywords (top matter com.), 297 rules for using, 280 keywords (bibl. com.), 425 \keywordsname (redef. name), 376 \kill (tabbing com.), 141, 142 Knuth, Donald E., xxix, 363, 521–523, 525, 528, 539, 553, 559 koi8u (font encoding), 548 L L, slashed (Ł, ł), 81, 515 \label (crossref. com.), 28, 29, 31, 50, 156, 215, 231, 252, 255, 328 and simple indexes, 267 arguments of, 29, 257 assigning counters to, 402 in table environments, 259 placement of commands, 486 rules for, 257 showing in margin, 257, 313 with \\, 215 labels for \bibitem commands, 263 for bibliographic items, 48, 121 for equations, 255 for groups of equations, 204 in list environments, 121, 121 of items in a list environment, 408, 409, 410 setting with \label, 255 \labelsep (length com.), 407, 408 \labelwidth (length com.), 407, 408 \Lambda (Λ Greek char.), 502
Index \lambda (λ Greek char.), 502 Lamport, Leslie, xxix, 403, 522, 523, 529, 552, 559 \land (∧ math op.), 506 landscape (doc. class opt.), 307 \langle (h math delimiter), 167, 168, 509 language (B IBTEX database field), 425 \LARGE (font size com.), 93, 93, 517 \Large (font size com.), 93, 93, 94, 517 large math delimiters in array subsidiary math environment, 313 operators, 25, 26, 161, 175, 174–176, 511 limits of, 174, 299 sizing of math delimiters with, 169 symbols, 117, 227–230 \large (font size com.), 93, 93, 517 \larger (font size com.), 94 LATEX, 521, 525 and the Internet, 537–542 counters, 399 customizing, 363–417 distribution, 78, 86, 310–313, 489, 525, 528, 529, 532, 534, 539, 540 components of, 303–308 directories, 257, 310–311, 312, 540, 547, 553 on CTAN, 540 document classes, see under document classes document structure, 247–270, 467 documentation for, 311, 552 packages, 312 documents, 247–270 printing and viewing, 530 putting on the Internet, 317–323 files created by, 531–534 files, portability of, 9 font substitution, 94, 318, 484 formats, 529 history of, 521–525 implementations, xxvi, 3, 5, 311, 318, 326, 525, 537 discussion groups, 542 for Mac, 3, 326, 489, 495–499 PC, 3 for PC, 326, 489–494 productivity tools, 5, 494, 498 technical support for, 541–542 UNIX , 3, 442, 461, 489 inner workings of, 528–534 layers of, 528–529 localization of, 79, 547–548 numbers stored by, 80
Index
LA–LE omissions in discussion of, 551–552 overview of, 521–536 packages, see under packages release dates of, 86, 251, 532 resources on the Internet, 542 source files, see under source files spacing in text, 11 structure of, 525, 526 updates to, 532 using, 525, 527 versions of, 532–534 specifying, in documents, 86, 250, 386, 532 writing books with, 465–487 latex (LATEX distr. directory), 499 \LaTeX (LATEX logo com.), 8, 80 LATEX 2.09, 386, 522–524 font names, 313 twoletter font commands, 94, 94 LATEX 2ε , see LATEX latex.ltx, 311 LATEX3, 523, 524, 541 \LaTeXe (LATEX 2ε logo com.), 80, 524 latexsym (pack.), 37, 180, 249, 294, 300, 311 layers, of LATEX and TEX, 528–529 layout of a list, 408 of Computer Modern typewriter font, 78, 78 of pages, 268–270, 313, 525 \layout (pagelayout diagram com.), 268 layout (pack.), 268, 313 \lbrace ({ math delimiter), 167, 509 \lbrack ([ math delimiter), 167, 509 \lceil (d math delimiter), 167, 509 \ldots ( . . . ) in math, 160, 186 in text, 24 \le (≤ binary rel.), 503 \leadsto (; math arrow), 507 \left (math delim.), 167–168, 170, 184, 235 blank, 168, 182 must be balanced, 168, 217 left double quote (“) text symbol, 75, 81, 518 typing, 11 left justification of text, 8 left single quote (‘) key, 9, 11, 62, 75 text symbol, 75, 519 \left( ( ( math delimiter), 24, 26, 166, 168, 184, 217, 235 \left. (blank math delim.), 168, 182 \left< ( h math delimiter), 168 \left[ ( [ math delimiter), 24, 166, 182
587 \Leftarrow (⇐ math arrow), 507 \leftarrow (← math arrow), 364, 374, 507 \leftarrowtail ( math arrow), 507 \leftharpoondown () math arrow), 507 \leftharpoonup (( math arrow), 507 \leftleftarrows (⇔ math arrow), 507 \leftmargin (length com.), 407, 408, 411 \Leftrightarrow (⇔ math arrow), 507 \leftrightarrow (↔ math arrow), 507 \leftrightarrows ( math arrow), 507 \leftrightharpoons ( math arrow), 507 \leftrightsquigarrow (! math arrow), 507 \leftroot (rootadjustment com.), 162 \leftthreetimes (h math op.), 506 \left (  math delimiter), 24, 26, 168, 180, 189, 191 legacy document classes, 54, 118, 249, 254, 263, 273, 302, 305, 303–313, 529 documents, 319 environments, 123, 143–145, 407 quotation, 144 quote, 143 verse, 144 legacyarticle.tex (sample file), 304, 305 legalpaper (doc. class opt.), 85, 298, 306 Lehman, Philipp, 447 lemmas, see proclamations length commands, 108, 270, 398, 403–406 and boxes, 405 arithmetic with, 312, 405 \baselineskip, 95, 101 \baselinestretch, 99 defining new, 404 \depth, 108, 108, 111 \fboxrule, 110 \fboxsep, 110 \hangafter, 100 \hangindent, 99 \height, 108, 108, 111 in list environments, 407 \itemindent, 407, 408 \itemsep, 407, 413 \labelsep, 407, 408 \labelwidth, 407, 408 \leftmargin, 407, 408, 411 \listparindent, 407, 408, 414 \marginparpush, 404 \marginparwidth, 113 \medskipamount, 377, 406 \multlinegap, 213 \oddsidemargin, 270 \overfullrule, 96 \parindent, 404, 414
588
LE–LI length commands (continued) \parsep, 407, 408, 413, 414 \parskip, 404, 407, 408, 414, 481 \partopsep, 407 \rightmargin, 407, 408, 411 setting, 99, 100, 113, 404–406 \textwidth, 270, 398, 404 \topsep, 407, 408 \totalheight, 108, 108, 110, 111 \width, 108, 108, 111 lengths, rubber, see rubber lengths \leq (≤ binary rel.), 503 leqno (doc. class opt.), 298, 307 \leqq (5 binary rel.), 504 \leqslant (6 binary rel.), 504 less than ( ( i math delimiter), 168 \right] ( ] math delimiter), 24, 166, 182 \Rightarrow (⇒ math arrow), 507 \rightarrow (→ math arrow), 507 \rightarrowtail ( math arrow), 507 \rightharpoondown (+ math arrow), 507 \rightharpoonup (* math arrow), 507 \rightleftarrows ( math arrow), 507 \rightleftharpoons ( math arrow), 507 \rightmargin (length com.), 407, 408, 411 \rightrightarrows (⇒ math arrow), 507 \rightsquigarrow ( math arrow), 507
Index \rightthreetimes (i math op.), 506 \right (  math delimiter), 24, 26, 168, 180, 189, 191 ring (˚ text accent), 79, 516 ring A (˚ A), 81, 515 ring a (˚ a), 81, 515 \risingdotseq ( : binary rel.), 504 Ritter, R. M., 553 \rm (obs. LATEX 2.09 font com.), 94 \rmfamily (font shape com. dec.), 90, 123, 516 rmfamily (font shape env.), 123 \Roman (uppercase romannumeral counterstyle com.), 401 roman document font family, 89, 89, 544 font shape, 28, 88, 90, 123, 301, 516 math, 173, 196, 196 \roman (lowercase romannumeral counterstyle com.), 123, 401 roots, 27, 161–162 nth, 27, 162 placement of, 162 square, 27, 161 using struts with, 200 Rose, Kristoffer H., 242 row separator (\\), 25, 30, 31, 33, 44, 46, 98, 124, 135, 141, 142, 145, 176, 211, 213, 215, 216, 227, 231, 242, 273–278, 297, 304, 309 Rowley, Chris, xxxiii, 523, 524, 559 \Rrightarrow (V math arrow), 507 \Rsh ( math arrow), 507 \rtimes (o math op.), 506 rubber lengths, 378, 406, 407, 530 \rule (box com.), 113, 114, 140 rules (lines), filling lines with, see also lines (rules), 106, 113 running BibTEX, 437–446 MakeIndex, 459–461 running heads, 56, 252, 268–270 author’s name in, 276 fragile commands in, 74 title in, 274, 297 \rVert ( k math delimiter), 168 \rvert (  math delimiter), 168 S \S (§) math symbol, 508 section text symbol, 81, 518 sampartref.tex (sample file), 320, 323 sampart.tex (sample file), 125, 129, 131, 230, 266, 272, 273, 285–293, 371, 392, 437, 536
Index
SA–SC source file, 290–293 typeset, 286–288 sampartb.aux (sample aux. file), 441, 442 sampartb.bbl (sample bibl. file), 441–443 sampartb.bib (sample bibl. file), 437, 437–443 sampartb.blg (sample BibTEX log file), 441, 442 sampartb.tex (sample file), 421, 437, 439–441, 443 sampartu.tex (sample file), 376, 383, 392–398, 535 sample files amsart.tpl, 294 amsproc.template, 297 babybeamer1.pdf, 333–335 babybeamer2.tex, 334 babybeamer3.tex, 338 babybeamer4.tex, 343 babybeamer5.tex, 341 babybeamer6.tex, 343 babybeamer6block.tex, 343 babybeamer7.tex, 343, 345–347 babybeamer8.tex, 347 babybeamer9.tex, 349 babybeamer10.tex, 359 beamerstructure.tex, 352, 354 bibl.tpl, 47 cleardoublepage.sty, 101 fonttbl.tex, 78, 134 formula.tex, 164–166 gg.tex, 44–48 gg2.tex, 45–46 ggamsart.tpl, 297 inbibl.tpl, 263 intrart.tex, 35–43, 53, 55, 326, 331 typeset, 41–43 intrarti.tex, 450–452, 460, 459–461 intropres.tex, 53, 55, 57–58 legacyarticle.tex, 304, 305 letter.tex, 308–310 math.tex, 17–19 mathb.tex, 19–22 myams.tpl, 294–297 newlattice.sty, 311, 379, 383, 386–392 note1.tex, 9–11 note1b.tex, 12–13 note2.tex, 13–15 noteslug.tex, 13 obtaining, 4 of presentations, 54, 326–331 products.eps (illus. file), 39
603 products.pdf (illus. file), 39, 54, 327 quickbeamer.pdf, 326 quickbeamer.tex, 326–328 quickbeamer1.pdf, 328 quickbeamer1.tex, 328 quickbeamer2.pdf, 331 quickbeamer2.tex, 331 sampartref.tex, 320, 323 sampart.tex, 125, 129, 131, 230, 266, 272, 273, 285–293, 371, 392, 437, 536 source file, 290–293 typeset, 286–288 sampartb.bib, 437, 437–443 sampartb.tex, 421, 437, 439–441, 443 sampartu.tex, 376, 383, 392–398, 535 sample.cls (doc. class), 10, 11 SymbolTables.pdf, 4, 9, 15, 19 template.bib, 423, 437, 446 topmat.tpl, 282 sample.cls, 11 sample.cls (doc. class), 10 samples font, 313 of bibliographies, 47, 261–263, 423, 437, 437–438, 446 of command (style) files, 311, 379, 383, 386–398 of indexes, 452, 450–452 samples (folder), 4, 9, 10, 12–15, 18–20, 35, 39, 44, 45, 47, 48, 78, 101, 134, 230, 263, 282, 285, 294, 297, 304, 308, 320, 326–328, 331, 333, 339, 341, 343, 354, 358, 359, 376, 386, 392, 423, 437, 438, 450, 468, 496, 535, 548 creating, 4 sans serif document font family, 89, 94, 544 font shape, 55, 89, 90, 123, 196, 196, 512, 516 Helvetica, 89, 544 \sb (math subscript com.), 159 \sc (obs. LATEX 2.09 font com.), 94 scaling graphics, 54 in figures, 259 in FoilTEX, 54 Sch¨ opf, Rainer, 523, 559, 560 Schandl, Bernd, 123 scharfes s (ß, SS), 81, 515 Schleyer, Antoinette Tingley, 217, 553, 560 school (BibTEX database field), 425 Schr¨ oder, Martin, 523
604
SC–SH scope delimited by braces, 71–73 of \allowdisplaybreaks commands, 242 of command declarations, 124, 367–368, 382 of commands, 71–73, 90, 139 global and local, 74 setting, with environments, 96, 213 \scriptscriptstyle (math font size com.), 199, 204 \scriptsize (font size com.), 93, 93, 517 \scriptstyle (math font size com.), 199, 204 \scshape (font shape com. dec.), 90, 91, 123, 516 scshape (font shape env.), 123 \searrow ( & math arrow), 507 \sec (sec math op.), 172, 510 secnumdepth (counter), 403 \section (struct. com.), 46, 55, 252–255, 331, 350, 467 optional argument of, 253 section (counter), 398, 399 section (§ text sym.), 81, 518 \section* (struct. com.), 46, 253, 467 sectioning commands, see also structural commands optional arguments of, 253 provided by amsart doc. class, 254 provided by article doc. class, 254 provided by book doc. classes, 466–467 syntax of, 253 of articles, 46, 252–255 of books, 466–467, 479 of documents, 11, 35–41, 46, 48, 49, 51, 52, 66, 68, 164, 248, 251, 252–255, 261, 267, 285, 304, 308–310, 406, 437, 442, 449, 459, 466–467, 522, 536 of presentations, 55, 328, 331 \sectionname (redef. name) and hyperref package, 323 sections, numbering of, 252 in books, 466 of equations within, 156, 250 of proclamations within, 127 \see (indexing crossref. com.), 457 \seename (redef. name), 376 \selectfont (font selection com.), 95 selecting
Index files to include, 260, 400, 477, 478, 479, 496 fonts, 88–95 semicolon (;), 9, 62 sentences terminating, 11, 64–67 with periods, 279 series (BibTEX database field), 425 series commands, for fonts, see under font commands and font command declarations serif (font shape), see also under roman, 89, 89 Seroul, Raymond, 553, 560 \setbeamercolor (beamer com.), 350 \setbeamertemplate (beamer com.), 355 \setcounter (countersetting com.), 74, 233, 399, 402 \setlength (length setting com.), 99, 100, 113, 404 setlength environment, 96, 213 \setminus ( \ math op.), 506 setspace (pack.), 99 \settodepth (measurement com.), 406 \settoheight (measurement com.), 406 \settowidth (measurement com.), 406 \sffamily (font shape com. dec.), 90, 94, 123, 516 sffamily (font shape env.), 123 sgml (Standard Generalized Markup Language), 523 shape commands, for fonts, see under font commands and font command declarations \sharp (] math sym.), 508 short arguments, of userdefined commands, 373–374 arguments, of userdefined environments, 385 commands, 73, 90, 273, 284 pages, warnings about, 530 shorthand, see under userdefined commands \shortmid (p binary rel.), 504 \shortparallel (q binary rel.), 504 \shoveleft (align. com.), 214 \shoveright (align. com.), 214 \show (commandexamination com.), 377, 377, 378, 534 \showhyphens (hyph.disp. com.), 84 showidx (pack.), 311, 453 showkeys (pack.), 257, 313 options, 257
Index
SH–SP \showthe (valueexamination com.), 113, 377, 378, 534 shrinkable lengths, 56, 101, 102, 377, 406, 407, 413, 530 sidesetting, math symbols, 194–195 sidebars, in beamer presentations, 328, 331, 343 \sideset (math com.), 194–195 \Sigma (Σ Greek char.), 502 \sigma (σ Greek char.), 502 \signature (letter com.), 309 \sim (∼ binary rel.), 503 \simeq (' binary rel.), 503 simple alignment, of formulas, 31, 30–31 \sin (sin math op.), 26, 170, 172, 196, 510 single guillemet, 519 quote, 11 keys, 9, 23, 62, 75 text symbols, 519 single quotation, 519 singlelined boxes, see under boxes \sinh (sinh math op.), 172, 510 size of files pdf, 318 PostScript, 317, 319 of fonts, 55, 89, 93, 93–94, 123, 517 commands for, see under font commands document class options for, 297, 306 in math, 199 of hyperlinks in pdf files, 322 of math delimiters, 168–170, 184 of paper, document class options for, 298, 306 \sl (obs. LATEX 2.09 font com.), 94, 95 slanted (font shape), 88, 90, 91, 94, 95, 123, 156, 516 and italic correction, 92 context dependence of, 91 slashed L’s and O’s (l, L, ø, Ø), 81, 515 slides, see under presentations slides (obs. doc. class), 303, 529 \slshape (font shape com. dec.), 90, 91, 92, 94, 95, 123, 516 slshape (font shape env.), 123 slugs, 13, 96, 249, 300, 307 \SMALL (font size com.), 93, 93, 517 \Small (font size com.), 93, 93, 517 \small (font size com.), 93, 93, 517 small caps (font shape), 88, 90, 91, 123, 516
605 for abbreviations and acronyms, 91, 480 \smaller (font size com.), 94 \smallfrown (a binary rel.), 504 \smallint (∫ math sym.), 508 smallmatrix (subsid. math env.), 235 \smallsetminus (r math op.), 506 \smallskip (spacing com.), 104 \smallsmile (` binary rel.), 504 \smash (spacing com.), 115, 201 optional argument of, 201 \smile (^ binary rel.), 503 Snow, Wynter, 553, 560 Solaris, see unix solid boxes, 113–114, see also struts solution (beamer distr. directory), 358 somedefs (pack.), 313 sort keys, 455–456 case sensitivity of, 462 sorting, of index entries, 455–456 source files, 4–5, 5, 7–8, 9, 18, 35, 51, 52, 64, 82, 87, 145, 271, 415, 447, 461, 473, 496, 525 errors in, 19–22, 491, 498, 534 naming of, 4, 44 platform independence of, 3, 9 putting on the Internet, 317–323 disadvantages of, 317 readability of, 19, 65, 154, 166, 363–367, 379, 415 structure of, see also document style parameters, see styles, document parameters for typesetting, 4, 9, 12, 20, 28, 63, 74, 529, 534 white space in, 19 sources, for fonts, 546 \sp (math superscript com.), 159, 183 space bar, 9, 11, 53, 63, 325, 334 spaces at the beginning of a line, 64, 103 at the end of a line, 64, 65, 84 blue, see also tie, unbreakable spaces, nonbreakable spaces, 76 consist of glue, 530 expanding, 106 filling lines with, 106 in arguments of commands, 85 \bibtem, 266 \cite, 264 \index, 462 in command definitions, 367 in command names, 69 in delimited commands, 379–380 in math environments, 153
606
SP–SQ spaces (continued) in tabular environments, 135 in userdefined commands, 313 in \verb commands, 148 in verbatim environments, 146 interword (\ ), 11, 66, 70, 102, 190, 520 multiple, act as one, 19, 64, 71 separating words with, 11, 64, 383, 427 suppressing, 383, 458 terminating commands with, 19, 69, 154 typed for readability, 22, 65, 135, 367 unbreakable (~ tie), 11, 28, 63, 66, 75, 76, 99, 520 absorb spaces, 76 in BibTEX databases, 427 with crossreferences, 257 visible ( ), 11, 63 and \verb* command, 148 spacing between characters, 530 between dots with \hdotsfor commands, 233 between math symbols, 187–192 commands, see spacing commands determined by document classes, 102 horizontal adjusting, 188 in math, 19, 154–156, 187–192, 233, 236 in text, 11, 102–104, 114, 126, 414, 513, 520 interword, 11, 64–67, 70, 76, 102, 383, 427, 520 preventing removal of, 103 in text, 102–106 intercolumn, 209, 220, 220 in aligned math environments, 224, 228 in tabular environments, 136 interline, 242, 370, 482 adjusting, 95, 98–99, 104, 114 double, 99 rules and delimited commands, 379–380 for commands, 69, 154 in arguments of \index commands, 462 in math, 19, 154–156, 187–192, 200–201 in text, 11, 19, 64–67, 102–106 in \text commands, 154
Index vertical adding to table of contents, 474 adjusting, 56, 139, 481 adjusting with the setspace package, 99 in boxes, 115 in indexes, 267, 449 in math, 200–201 in text, 15, 104–105, 201, 377, 474, 482, 486 preventing removal of, 105 using the setspace package, 99 spacing commands avoiding direct use of, 481 horizontal, see horizontal spacing vertical, see vertical spacing \spadesuit (♠ math sym.), 508 \spbreve (˘ math accent), 177, 177, 512 \spcheck ( ∨ math accent), 177, 183, 512 \spdddot ( ... math accent), 177, 512 \spddot ( .. math accent), 177, 177, 512 \spdot ( . math accent), 177, 512 special braces, 18, 49, 72, 73, 152, 153, 157 balancing, 72 characters, 15, 63, 76, 77, 77, 78, 515–519 in \index commands, 457 keys, 9, 17, 63, 76 spelling checkers, 48, 495 \sphat ( b math accent), 177, 512 \sphericalangle (^ math sym.), 508 Spivak, Michael D., 522, 545 split (subsid. math align. env.), 208, 211, 230–231, 298 and \allowdisplaybreaks commands, 243 numbering of equations within, 230–231 rules for, 231 splitting documents into multiple files, 476–479 formulas across lines, 212–214 \sptilde (∼ math accent), 177, 512 \sqcap (u math op.), 506 \sqcup (t math op.), 506 √ \sqrt ( x math com.), 19, 27, 161 optional argument of, 70 \sqsubset (< binary rel.), 503 \sqsubseteq (v binary rel.), 503 \sqsupset (= binary rel.), 503 \sqsupseteq (w binary rel.), 503 \square ( math sym.), 508 square roots, 27, 161
Index
SY–TE \SS (SS), 81, 515 \ss (ß), 81, 515 stacking math symbols, 192–193 Standard Generalized Markup Language (sgml), 523 \star (? math op.), 506 starred commands and environments, see under *form and specific commands and environments start of chapters, document class options for, 468 \stepcounter (counterincr. com.), 402 sterling (£ text sym.), 519 \stop (interactive control com.), 535 stretchable horizontal braces, 178–179 lengths, 102, 378, 406, 407, 530 lines, 179 math arrows, 179, 240–242 math delimiters, 167–168 math symbols, 178–180 vertical spacing, in displayed text environments, 406 STRING (bibl. entry type), 436 structural commands \appendix, 255, 467 \backmatter, 467 \chapter, 253, 255, 304, 466, 466, 467 \frontmatter, 467 hierarchy of, 252, 466 in presentations, 350 \mainmatter, 467 optional arguments of, 253 \paragraph, 252, 466 \paragraph*, 253 \part, 253, 304, 466 provided by amsart doc. class, 254 provided by article doc. class, 254 provided by book doc. classes, 466–467 \section, 46, 252–255, 467 \section*, 46, 253, 467 \subparagraph, 252, 466 \subsection, 46, 252, 466, 467 \subsection*, 46, 467 \subsubsection, 46, 252, 466 \subsubsection*, 46 syntax of, 253 structures, theoremlike, see proclamations struts, 114, 139, 200 style (sty) files, see also packages, 250, 366, 386–392 commands in, 386 terminating, 392
607 style files, BibTEX, see bst styles bibliographical, see also bst, 266 document parameters for, 552 for counters, 87, 123, 382, 401, 400–401, 411 of pages, 268–270 of proclamations, 129–131, 524 of tables, 140 text environments, 123 subarray (subsid. math env.), 161, 176 subentries, in indexes, 267, 449, 453, 454, 456 subequations (disp. math env.), 203, 204, 219, 401 subformulas indentation of, 217 rules for, 215–217, 223–224 shorthand commands for, 364–366, 372–373 \subitem (index com.), 267, 449 \subjclass (top matter com.) optional arguments of, 279 rules for using, 279 \subjclassname (redef. name), 376 subject classifications, AMS, 279 \subparagraph (struct. com.), 252, 466 subparagraph (counter), 399 subscripts, math, 23, 25, 158–159, 161, 182 as limits, 25, 161, 171–176, 182, 299 as multiline limits, 176 double, 199 font size of, 199 with horizontal braces, 178 \subsection (struct. com.), 46, 252, 350, 466, 467 subsection (counter), 399 \subsection* (struct. com.), 46, 467 \subsectionname (hyperref redef. name), 323 \Subset (b binary rel.), 504 \subset (⊂ binary rel.), 503 \subseteq (⊆ binary rel.), 503 \subseteqq (j binary rel.), 504 \subsetneq (( neg. binary rel.), 505 \subsetneqq ($ neg. binary rel.), 505 subsidiary math environments aligned, 211, 227–230, 243 alignedat, 227–230, 243 and \allowdisplaybreaks commands, 243 array, 208, 210, 229, 232, 236–238, 312, 313, 523, 551, 552 Bmatrix, 235 bmatrix, 235
608
SU–SY subsidiary math environments (continued) cases, 33, 208, 210, 239 CD, 240–242 for aligning formulas, 208, 209, 227–231, 243, 298 gathered, 211, 227–230, 243 matrix, 25–26, 208, 210, 231, 232–238 numbering of, 230, 231 page breaks in, 242 pmatrix, 25–26, 235 smallmatrix, 235 split, 208, 211, 230–231, 243, 298 subarray, 161, 176 Vmatrix, 183, 235 vmatrix, 25, 26, 235 \substack (math com.), 161, 176, 176 substitution, of fonts, 94, 318, 484 subsubentries, in indexes, 267, 449, 454, 456 \subsubitem (index com.), 267, 449 \subsubsection (struct. com.), 46, 252, 466 subsubsection (counter), 399 \subsubsection* (struct. com.), 46 \subsubsectionname (redef. name), 322 subtraction, 22, 63, 157 \succ ( binary rel.), 503 \succapprox (v binary rel.), 504 \succcurlyeq (< binary rel.), 504 \succeq ( binary rel.), 503 \succnapprox ( binary rel.), 505 \succneqq ( binary rel.), 505 \succnsim ( binary rel.), 505 \succsim P (% binary rel.), 504 \sum ( large math op.), 26, 175, 198, 511 with primes, 194 sumlimits (opt. of amsmath pack.), 299 sums, 26, 27, 174 \sup (sup math op.), 172, 510 superscripts, 23, 25, 158–159, 161, 177, 182 as limits, 25, 161, 171, 174–176, 182 as multiline limits, 176 in math as limits, 299 font size of, 199 with horizontal braces, 178 in text, 311, 519 support, technical, see technical support \suppressfloats (float com.), 260 suppressing creation of auxiliary files, 486, 531 hyperlinks in pdf files, 322
Index indents (\noindent com.), 99, 484 italic correction, 92 ligatures, 79, 181, 365 line breaks (\nolinebreak com.), 99 loading of the amsmath package, 300 numbering, 211, 215 of equations, 30, 31, 157, 201, 211, 213, 215, 218, 231, 270 of first page of documents, 270 of proclamations, 129, 130, 382 of structural units, 46, 253, 467 Overfull \hbox warnings, 96 page breaks (\nopagebreak com.), 100 placement of floats, 260 q.e.d. symbol, 132 spaces, 383, 458 \Supset (c binary rel.), 504 \supset (⊃ binary rel.), 503 \supseteq (⊇ binary rel.), 503 \supseteqq (k binary rel.), 504 \supsetneq () binary rel.), 505 \supsetneqq (% binary rel.), 505 √ \surd ( math sym.), 508 Sutor, Robert, 558 Swanson, Ellen, 217, 553, 560 \swapnumbers (theoremstyle com.), 131 \swarrow ( . math arrow), 507 \symbol (symbol com.), 77, 311 symbol alphabets, math, 197 blackboard bold, 197 calligraphic, 197 Euler Fraktur, 197 Greek, 197 symbolic references, showing in margins, 257, 313 symbolic referencing, 28, see also crossreferencing, 29, 255–258, 531 symbols as footnote indicators, 87 math, 19, 20, 63, 131, 177, 182, 240, 249, 294, 311, 501–513 alphabets, 197 and delimiters, 24, 166, 168 bold, 197–199, 301, 312 building new, 183, 192–195, 372 classification of, 188–189, 195 declaring types of, 195 end of proof, 113, 374 in text, 365 large, 227–229 negated, 194–195 shorthand commands for, 364 sidesetting, 194–195
Index
SY–TE sizes of, 199 spacing of, 187–192 stacking, 192–193 stretchable, 178–180 suppressing, 131 text, 76, 77, 80, 278, 516, 518, 519 SymbolTables.pdf (sample file), 4, 9, 15, 19 synchronization (inverse search), 5, 494, 498 systems of equations, 225–226 systems, operating, see Mac, PC, or unix T \t ( tie text accent), 79, 516 T1 font encoding, 82, 519, 547 tab character, 63, 64, 68, 102 Tab key, 9, 11, 63 Tab key, 9, 11, 63 tabbing (display text env.), 141–143, 551, 552 table (counter), 399 table (float env.), 258–259 optional arguments of, 260, 486 placement of, 484, 486 table* (float env.), 259 \tablename (redef. name), 376 and hyperref package, 323 \tableofcontents (frontmatter com.), 328, 473, 532 tables, 133–140, 141, 237, 259, 260, 313, 523, 551, 552 captions in, 134, 259, 476 fragile commands in, 74 commands for, 135, 137–140, 259, 476 crossreferencing of, 134 designing, 140 doublecolumn, 259 float controls, 134 forcing typesetting of, 101 hyphenation, 528 intersection of lines in, 140, 313 lists of, see under lists multiline entries in, 138 multipage, 313 numbering of, 258 placement of, 484, 486 style of, 140 typeset inline, 133 tables of contents, 252, 321, 473–475 adding a line to, 473–474 adding vertical spacing to, 474
609 auxiliary file (toc), 473, 487, 532 commands, 473, 474 depth of, 403 fragile commands in, 74, 475 in frames, 328 in presentations, 358 typesetting, 473–475 tabs, setting, 141 tabular (table env.), 78, 133–140, 237, 259, 312, 313, 523, 551, 552 arguments of, 135 optional, 135 breaking lines in, 135 columnformatting in, 136 horizontal lines in, 135, 137 intercolumn space in, 136 \parbox in, 111 rules for, 135 vertical lines in, 135 width of columns in, 136 tabularx (pack.), 313 \tag, 29, 201, 202, 211, 213, 215, 218, 230, 231 \tag*, 202 tags (names for equations), 29, 201–204, 211, 213, 215, 218, 230, 231 toporbottom placement of, 298 \tan (tan math op.), 172, 510 \tanh (tanh math op.), 172, 510 Tantau, Till, 325, 359, 552, 560 \tau (τ Greek char.), 502 \tbinom (inlinestyle binomial com.), 160 tbtags (doc. class opt.), 298 technical reports, in BibTEX database files, 434–435 technical support on the Internet, 542 provided by TEX users groups, 541 provided by the AMS, 542 TECHREPORT (bibl. entry type), 425, 434–435 template.bib (sample bibl. file), 423, 437, 446 templates, 271, 306 bibliographical, 47, 261, 423, 437, 446 customizing for AMS document classes, 294–297 for articles, 44–45 ggamsart.tpl, 297 myams.tpl, 294–297 personalized making readonly, 297
610
TE–TE terminating captions, 476 commands, 14, 19, 69, 70, 80, 147, 154 environments, 68 fields in BibTEX database entries, 445 files, 392, 478 lines, 85, 159, 367 paragraphs, 64, 99 sentences, 11, 64–67, 279 TEX, 521, 525 commands, 378, 380 to avoid using, 481 commands in LATEX, 482 distribution, 528 extensibility of, 525 files, portability of, 9 history of, 521–525 implementations, 489 inner workings of, 528 inputs folders, 540, 544–545 layers of, 528–529 omissions in discussion of, 552 Plain, 528 programming in, 525, 553 resources on the Internet, 542 source files, see under source files structure of, 528 users groups, see TEX users groups and tug tex (LATEX distr. directory), 499 \TeX (TEX com.), 80 TEX users groups, see also tug, 537–542, 547 international, 541 tex4ht (opt. of hyperref pack.), 320 TEX Live (tug dvd/cd), xxvi, 489, 490, 495, 541 texmf (LATEX distr. directory), 499 TeXShop (LATEX front end for Mac), xxvi, 9, 11, 53, 326, 496 customizing, 496 productivity tools in, 498 with BibTEX, 442 with MakeIndex, 461 text, 7 accents, 69, 79, 516 \\ in, 98 blocks, width of, 270 boxes, see text boxes centering, see centering text, center (text align. env.), and \centering editors, xxvi, 4, 5, 9, 18, 48, 55, 63, 84, 88, 365, 490, 526
Index framing, 109 in math mode, 19, 27, 83, 99, 107, 154, 162–163, 181, 196, 226–227, 241, 301 specifying fonts for, 163 in tables multicolumn, 137–140 multiline, 138 mode, 7 plain, 9 spacing in, 102–106 style commands, see text style commands symbols, see text symbols typing, 7–15, 61–115 userdefined commands for, 366 using math symbols in, 365 white space in, 102–106 \text (box com.), xxvi, 27, 33, 83, 99, 107, 162–163, 181, 206, 241, 301 spacing rules in, 154 text boxes, 107–115 alignment of contents in, 107, 111 behave as single characters, 107 commands for \framebox, 109 \makebox, 107, 108, 109, 111, 113 \mbox, xxvi, 107 \parbox, 107, 109, 110, 111, 138, 405 \raisebox, 115 \rule, 113, 114, 140 \text, xxvi, 27, 33, 83, 99, 107, 154, 162–163, 181, 206, 241, 301 environments for minipage, 107, 109, 112, 383, 384 framed, 109–110 measurements of, 405 measuring, see measurements single line, 107–109, 163 vertical alignment of, 111, 115 text environments, 117–149 and stretchable vertical space, 406 blank lines in, 118 displayed, see displayed text environments for abstracts, see abstract for alignment, 8, 68, 106, 124, 134, 227, 407, 414 for bibliographies, see thebibliography for boxes, see under text boxes for comments, see comment for documents, see document for figures, see figures
Index
TE–TH for indexes, see theindex for letters, see letter for lists, see list text environments for proofs, see proof for styles, see under text style for tables, see tables rules for, 118 text mode, 62 text style commands avoiding direct use of, 480 \emph, 8, 14, 68, 90, 91, 337, 516 \textbf, 14, 73, 90, 93, 163, 301, 337, 516 \textit, 90, 91, 337, 516 \textmd, 90, 93, 516 \textnormal, 90, 90, 163, 516 \textrm, 90, 301, 516 \textsc, 90, 91, 516 \textsf, 90, 516 \textsl, 90, 91, 516 \texttt, 14, 76, 90, 516 \textup, 90, 91, 516 text style environments bfseries, 123 em, 123 itshape, 123 rmfamily, 123 scshape, 123 sffamily, 123 slshape, 123 ttfamily, 123 upshape, 123 text symbols, 77, 81 commands, 76, 77, 80, 278, 516, 518, 519 typing, 74–80 \textasciicircum (ˆ circumflex), 81, 518 \textasciitilde ( ˜ tilde), 519 \textasteriskcentered (∗ asterisk), 77, 81, 518 \textbackslash ( \ backslash), 76–77, 81, 518 \textbar (  vertical bar), 77, 81, 518 \textbf (font weight com.), 14, 73, 90, 93, 163, 301, 337, 516 \textbullet (• bullet), 81, 518 a 80, 81, 518 \textcircled ( ), \textcolor (beamer com.), 349 \textcompwordmark (lig.suppr. com.), 79, 365, 415 \textemdash ( — em dash), 81, 518 \textendash ( – em dash), 81, 518 \textexclamdown (¡ exclamation mark), 62 \textgreater ( > greater than), 81, 518
611 \textit (font shape com.), 90, 91, 337, 516 \textless (< less than), 81, 518 \textmd (font weight com.), 90, 93, 516 \textnormal (font shape com.), 90, 90, 163, 516 \TextOrMath (text and math mode com.), 311 \textperiodcentered (· midpoint), 81, 518 \textquestiondown ( ¿ question mark), 81, 518 \textquotedbl (Eur. quot. mark), 519 \textquotedblleft ( “ left double quote), 81, 518 \textquotedblright ( ” right double quote), 81, 518 \textquoteleft ( ‘ left single quote), 519 \textquoteright ( ’ ) Eur. quot. mark, 519 right single quote, 519 R registered \textregistered ( trademark), 81, 518 \textrm (font shape com.), 90, 301, 516 \textsc (font shape com.), 90, 91, 516 \textsf (font shape com.), 90, 516 \textsl (font shape com.), 90, 91, 516 \textstyle (math font size com.), 199, 204 \textsubscript, 251, 311 \textsuperscript ( a ), 311, 519 \texttrademark ( TM trademark), 519 \texttt (font shape com.), 14, 76, 90, 516 \textup (font shape com.), 90, 91, 516 textures (opt. of hyperref pack.), 320 \textvisiblespace, 11, 63, 64, 519 \textwidth (length com.), 270, 398, 404 tfm (LATEX font metric file), 529 \tfrac ( x ), 158, 200 y \TH (Thorn Eur. char.), 519 \th (thorn Eur. char.), 519 \thanks (top matter com.), 44, 87, 279, 281, 296, 304 \the (value expansion com.), 80, 87, 113, 400 thebibliography (bibliography env.), 40, 47–48, 66, 261–267, 407, 437, 446 argument of, 263, 266 in bbl files, 442 \thechapter (the value of counter chapter), 401 \thefootnote (the value of counter footnote), 87, 88
612
TH–TO theindex (index env.), 267, 407, 449, 459 commands in, 267, 449 themes (beamer distr. directory), 356 themes, for beamer presentations, 328, 356–357 Berkeley, 328, 331, 332, 356 color options, 357 Warsaw, 331, 332, 356 theorem (proclamation env.), 37, 52, 54, 125, 128, 343, 384 body of, 46 logical design of, 52 visual design of, 52 theorem (pack.), 313 theoremlike structures, see proclamations \theoremname (hyperref redef. name), 323 \theoremstyle* (procl. com.), 37, 129–131 \theoremstyle (procl. com.), 129–131 \theparentequation (the value of counter parentequation), 401 \therefore ( ∴ binary rel.), 504 \thesection (the value of counter section), 401 theses, in BibTEX database files, 433–434 \thesubsection (the value of counter subsection), 401 \Theta (Θ Greek char.), 502 \theta (θ Greek char.), 502 \thickapprox (≈ binary rel.), 504 \thicksim (∼ binary rel.), 504 \thickspace (spacing com.), 190, 513, 520 \thinspace (spacing com.), 25, 67, 75, 188, 190, 190, 513, 520 thirdparty packages, see specific packages \thispagestyle (set page style), 270 thorn (Eur. char), 519 Thorup, Kresten K., 403 tie (~ spacing com.), 11, 28, 63, 66, 75, 76, 99, 520 absorbs spaces, 76 in BibTEX databases, 427 with crossreferences, 28, 257 tie ( text accent), 79, 516 TikZ, 552 tilde ( ˜ ) text accent, 79, 516 text symbol, 519 \tilde (˜ x math accent), 25, 177, 512 \time (time com.), 80
Index time commands \date, 44 \day, 80 \month, 80 \time, 80 \today, 14, 44, 70–71, 80, 148, 275, 304, 380, 548 \year, 80 Times, in LATEX documents, 544–545 \times (× math op.), 23, 158, 506 times (pack.), 544 times.sty (psnfss distr. file), 544 \Tiny (font size com.), 93, 93, 517 \tiny (font size com.), 93, 93, 517 \title (top matter com.), 37, 44, 106, 249, 297, 304, 331, 469 optional arguments of, 273, 274, 282, 297 title (BibTEX database field), 424–426, 433 title pages, 273, 535 commands for, 54 document class options for, 249, 299, 300, 307 environments for, 469 footnotes on, 87 of articles, 37, 249, 251 of books, 467, 469 of presentations, 55 titlepage (doc. class opt.), 249, 299, 307, 468 \titlepage (beamer com.), 328 titlepage (front matter env.), 469 titles in bibliographies rules for typing, 426–427 in running heads, 274, 297 of articles, 44, 273, 297 of structural commands, 252 fragile commands in, 74 \to (→ math arrow), 507 toc (table of contents files), 473–475, 487, 532 commands in, 475 tocdepth (counter), 403, 473 \today (time com.), 14, 44, 70–71, 80, 148, 275, 304, 380, 548 tokens, 529 tools (LATEX distr. directory), 311–313 packages in, 312–313 tools (pack.), 312 \top (> math sym.), 508 top matter, 36, 37–38, 249, 251, 272 article information, 37–38, 44–46 of AMS documents, 273–275
Index
US–VE commands, 37, 44, 46, 80, 87, 106, 270, 273, 273–281, 282, 284, 285, 296, 297, 331, 469 blank lines in, 273 examples of, 282–285 with multiple authors, 281 customizing templates, 44, 294 editing, 44–46 is documentclass specific, 251 of AMS documents, 273–285 AMS information, 279–281 author information, 275–279 errors with, 284–285 examples, 282–285 with multiple authors, 281 of books, 251 of presentations, 53–55, 326–327, 331 with multiple authors, 46 toporbottom, placement of tags, 298 topmat.tpl (sample file), 282 \topsep (length com.), 407, 408 \totalheight (length com.), 108, 108, 110, 111 trace (pack.), 313 R trademark text symbols (TM ), 81, 518, 519 trans (beamer doc. class opt.), 358 translation, of line ending characters, 63 \translator (top matter com.), 274 \triangle (4 math sym.), 508 \triangledown (O math sym.), 508 \triangleleft (/ math op.), 506 \trianglelefteq (E binary rel.), 504 \triangleq (, binary rel.), 504 \triangleright (. math op.), 506 \trianglerighteq (D binary rel.), 504 trivlist (list text env.), 414 \tt (obs. LATEX 2.09 font com.), 94 \ttfamily (font shape com. dec.), 90, 123, 142, 516 ttfamily (font shape env.), 123 tug (TEX Users Group), 489, 490, 495, 537, 541–542 TUGboat, 541, 553 twoletter font commands, see under font commands twocolumn (doc. class opt.), 85, 101, 259, 299, 307, 308 and footnotes, 313 \twocolumn (doublecolumn com.), 101 \twoheadleftarrow ( math arrow), 507 \twoheadrightarrow ( math arrow), 507 twoside (doc. class opt.), 85, 249, 299, 307, 308, 468
613 type (BibTEX database field), 425 type foundries, 544, 546 types of commands, 73–74 of math symbols, 188, 198 declaring, 195 typeset example articles intrart.tex, 41–43 sampart.tex, 285–288 files, 5, 5, 271, 443 dvi format, 318, 530 pdf format, 5, 259, 318, 326, 530 typesetting indexes, 459–461 source files, 4, 9, 12, 20, 28, 63, 74, 534 tables of contents, 473–475 with LATEX, 522, 526, 527, 529–534 typewriter style document font family, 89 font shape, 14, 76, 78, 90, 123, 142, 365, 516, 544 fonts, 88 in math, 196, 196, 512 typing bibliographic entries, 426–437 accents in, 427 names in, 426 rules for, 443–446 titles in, 426–427 books, 465–487 exercises, 476 footnotes, 87–88 math, 17–33, 151–206 poetry, 144–145 programs, 141–143, 145–147 quotations, 143–144 text, 7–15, 61–115 symbols, 74–80 typographical errors, 48 rules for abbreviations and acronyms, 91 initials, 67 U \u breve text accent ( ˘ ), 79, 516 uk tug (United Kingdom TEX Users Group), 541 Frequently Asked Questions (faq), 542 \ulcorner (p math delimiter), 167, 509 Umeki, Hideo, 270
614
UM–US umlaut ( ¨ text accent), 9, 15, 79, 516, 548 unbreakable hyphens, 83 spaces (~ tie), 11, 28, 63, 66, 75, 76, 99, 520 absorb spaces, 76 in BibTEX databases, 427 with crossreferences, 257 \underbrace (math sym.), 178 with a subscript, 178 underdot text accent (.), 79, 516 \underleftarrow ( x math arrow), 179 ← − \underleftrightarrow ( x math arrow), ← → 179 \underline ( x math line sym.), 179 \underrightarrow ( x math arrow), 179 − → underscore ( ), 77 text symbol, 81, 518 in email addresses, 278 underscore text accent ( ), 79, 516 ¯ \underset (symbolbuilding com.), 183, 192 Unicode, 554 units absolute, see under absolute dimensional, see under dimensional units logical, see under logical math, see mu relative, see under relative unix, 3, 9, 522, 554 ftp clients, 538 LATEX implementations, 3, 442, 461, 489 line ending characters, 9 \unlhd ( math op.), 506 unmarked footnotes, 44, 279 unnumbered environments, 215 equations, 30, 31, 157, 201, 202, 210, 211, 213, 215, 218, 231, 328 items, 119 math environments, 211 proclamations, 129, 130, 382 structural units, 46, 253, 467 unpacked (LATEX distr. directory), 311, 540 UNPUBLISHED (bibl. entry type), 435 \unrhd ( math op.), 506 \Uparrow (⇑) math arrow, 507 math delimiter, 167, 509 \uparrow (↑)
Index math arrow, 507 math delimiter, 167, 509 updates to AMSFonts and packages, 533 to LATEX, 532 \Updownarrow (m) math arrow, 507 math delimiter, 167, 509 \updownarrow (l) math arrow, 507 math delimiter, 167, 509 \upharpoonleft ( math arrow), 507 \upharpoonright ( math arrow), 507 \uplus (] math op.), 506 \upn (font shape com.), 156 uppercase counter styles letters (\alph), 401 roman numerals (\roman), 401 upref (pack.), 300, 301 upright (font shape), 28, 88, 89, 90, 91, 123, 156, 382, 412, 516 context dependence of, 91 \uproot (rootadjustment com.), 162 \upshape (font shape com. dec.), 90, 91, 123, 382, 412, 516 upshape (font shape env.), 123 \Upsilon (Υ Greek char.), 502 \upsilon (υ Greek char.), 502 \upuparrows ( math arrow), 507 \urcorner ( q math delimiter), 167, 509 \URL (bibl. com.), 425 url (Uniform Resource Locator), 278, see also hyperlinks \urladdr (top matter com.), 278, 279, 296, 323 \usecolortheme (beamer com.), 357 \usecounter (list counter com.), 411 \usefonttheme (beamer com.), 357 \usepackage (preamble com.), 35, 53, 70, 249, 294, 320, 326, 366, 386, 450, 534 can load multiple packages, 250 userdefined commands, 36, 364–380, 480 arguments of, 370–374 as shorthand, 364–374 as tokens, 529 command declarations in, 90 scope of, 367–368, 382 defining, 87, 99, 132, 139, 266, 267, 364, 366, 373–375, 400 in preamble, 250 delimited, 378–380 for environments, 365 for indexing, 371, 454 for subformulas, 371–373
Index
VE–$$ for text, 366 invoking, 366 multiple arguments of, 70 optional arguments of, 374 rules, 366, 367 short arguments, 373–374 with \def, 378 userdefined environments, 380–385, 479, 482, 529 defining, 381 for custom lists, 406–415 optional arguments of, 384 short arguments, 385 users groups, 537, 539, 541–542, 547 international, 541 \usetheme (beamer com.), 328, 356 unix, 479 V \v caron text accent ( ˇ ), 79, 516 \value (value of counter), 402 values, of counters, 87, 402 printing, 80, 113, 400 van Oostrum, Piet, 270, 537 \varDelta ( ∆ Greek char.), 502 \varepsilon (ε Greek char.), 502 \varGamma ( Γ Greek char.), 502 \varinjlim (lim math op.), 172, 510 −→ varioref (pack.), 258, 313 \varkappa (κ Greek char.), 502 \varLambda ( Λ Greek char.), 502 \varliminf (lim math op.), 172, 510 \varlimsup (lim math op.), 172 \varnothing (∅ math sym.), 508 \varOmega (Ω Greek char.), 502 \varPhi (Φ Greek char.), 502 \varphi (ϕ Greek char.), 502 \varPi (Π Greek char.), 502 \varpi ($ Greek char.), 502 \varprojlim (lim math op.), 172, 510 ←− \varpropto (∝ binary rel.), 504 \varPsi (Ψ Greek char.), 502 \varrho (% Greek char.), 502 \varSigma (Σ Greek char.), 502 \varsigma (ς Greek char.), 502 \varsubsetneq ( neg. binary rel.), 505 \varsubsetneqq (& neg. binary rel.), 505 \varsupsetneq (! binary rel.), 505 \varsupsetneqq (' binary rel.), 505 \varTheta (Θ Greek char.), 502 \vartheta (ϑ Greek char.), 502 \vartriangle (M math op.), 506 \vartriangleleft (C math op.), 506 \vartriangleright (B math op.), 506
615 \varUpsilon ( Υ Greek char.), 502 \varXi (Ξ Greek char.), 502 \Vdash ( binary rel.), 504 \vDash ( binary rel.), 504 \vdash ( ` binary rel.), 503 \vdots (vertical ellipsis), 161, 235 \vec (~ x math accent), 25, 177, 377, 512 \vee (∨ math op.), 506 \veebar (Y math op.), 506 \verb (inline verbatim com.), 147–149 in aligned math environments, 148 in argument of other commands, 148 spaces in, 148 \verb* (inline verbatim com.), 148 delimiters with, 147–149 verbatim (disp. text env.), 145–147, 313, 359, 523 blank lines in, 146 characters following, 146 simulating with \verb, 148 verbatim (pack.), 86, 87, 147, 313, 387 verse (disp. text env.), 144, 407 breaking lines in, 145 versions of AMS packages and AMSFonts, 300, 301, 524, 525, 533 of LATEX packages, 532–534 \Vert ( k math delimiter), 167, 509 \vert (  math delimiter), 167, 509 vertical alignment, of boxes, 111, 115 bar (  text symbol), 81, 518 ellipses, 161, 235 glue, 530 lines in CD environments, 241 in tabular environments, 135 math arrows, 241 vertical spacing adding after \\, 15 adding to table of contents, 474 adjusting, 56 with \arraystretch, 139 with struts, 139 adjusting with the setspace package, 99 between marginal notes, 404 between paragraphs, 404 commands \bigskip, 104 \enlargethispage, 101, 484, 487 \medskip, 104 \smallskip, 104 \smash, 115, 201, 201 to avoid using, 481
616
VE–WH vertical spacing commands (continued) \vfill (fill com.), 106 \vphantom, 105 \vspace, 104 \vspace*, 105 in boxes, 115 in indexes, 267, 449 in math mode, 200–201 in text, 15, 104–105, 377, 474, 482, 486 preventing removal of, 105 stretchable, 406 using the setspace package, 99 \vfill (space fill com.), 106 viewers dvi, 526, 530, 539 pdf, see under pdf viewing typeset files, 494, 530 dvi format, 318, 530 pdf format, 5, 259, 318, 530 virtex (TEX program), 528 visible spaces ( ), 11, 63 and \verb* command, 148 Visual Guide, to multiline math formulas, 207 visual design, 52–53, 247, 268–270, 358 markup, 535 guide, to multiline math formulas, 208 Vmatrix (subsid. math env.), 183, 235 vmatrix (subsid. math env.), 25, 26, 235 volume (BibTEX database field), 424, 425 von, in bibliographic entries, 426 \vphantom (spacing com.), 105 \vspace (spacing com.), 104 \vspace* (spacing com.), 105 vtex (opt. of hyperref pack.), 320 \Vvdash ( binary rel.), 504 W warning messages, see also error messages about hyphenation, 96 Abstract should precede \maketitle in AMS document classes, 251 Characters dropped after \end{verbatim}, 146 \end occurred inside a group at level x , 72 \end occurred when \xxx on line yy was incomplete, 478
Index generated by overlapping braces, 72 Label(s) may have changed, 256 line numbers in, 12 lines are too wide, 12–13, 96–97 No auxiliary output files, 531 Overfull \hbox, 12–13, 96–97, 530 suppressing, 96 recorded in log file, 96, 526 regarding font substitution, 94 Rerun to get crossreferences right, 256 Underfull \vbox has occurred while \output is active, 530 Warningmissing field in label , 444–446 when using \NeedsTeXFormat, 250, 386, 532 with BibTEX, 443–446 in log (blg) files, 442 Warsaw (beamer theme), 331, 332, 356 Web, see Internet \wedge (∧ math op.), 506 weight, of fonts, see under font environments, font commands and font declarations white space adjusting, 482 and binary operations and relations, 155 between marginal notes, 404 between paragraphs, 404 commands, avoiding direct use of, 481 detemined by document classes, 102 horizontal, 68, 102, 105, 142, 154, 189–191, 224, 225, 313, 368–369, 513, 520 in math mode, 186, 233, 236 in text mode, 11, 19, 64–67, 102–104, 106, 114, 126, 414 in arguments of commands, 85 in command definitions, 367 in command names, 68 in delimited commands, 379–380 in math, 19, 51, 104–105, 154–156, 187–192, 200–201 adjusting, 104 in source files, 154 in tabular environments, 135 in text, 11, 102–106, 383, 427, 513, 520 in \verb commands, 148 in verbatim environments, 146 suppressing, 383, 458 terminating commands with, 69
617
VE–$$
Index
vertical, 102 adding to table of contents, 474 adjusting, 481 in math mode, 104–105, 200–201 in text mode, 15, 101, 104–106, 201, 377, 482, 486 \widehat ( b x math accent), 177, 177, 184, 512 \widetilde ( e x math accent), 177, 177, 512 width of columns in tabular environments, 136 of fonts, see font width of text blocks, 270, 398, 404 of text boxes, 107–109, 405 \width (length com.), 108, 108, 111 Windows, see also PCs, see under Microsoft WinEdt (LATEX front end for PC), xxvi, 9, 11, 53, 490, 491, 492 installation, 490–491 productivity tools in, 494 with BibTEX, 442 with MakeIndex, 461 word processors, 11 words, are separated by spaces, 82 work (folder), 4, 10, 12, 14, 18, 20, 35, 39, 53, 164, 294, 320, 326–328, 331, 343, 438, 491 creating, 4 World Wide Web, see Internet \wp (℘ math sym.), 508 \wr (o math op.), 506 wrapping, of lines by email clients, 65 wrapping, of lines by text editors, 84 X xcb (exercise env.), 476 xcolor (pack.), 325, 348 options, 349 \Xi ( Ξ Greek char.), 502 \xi (ξ Greek char.), 502 \xleftarrow (stretchable math arrow), 179 xr (pack.), 258, 313 \xrightarrow (stretchable math arrow), 179 xspace (pack.), 313, 368–369 xypic (pack.), 242 Y Y&Y, 525, 543 \year (time com.), 80
year (BibTEX database field), 424, 425 X zero, specifying in arguments, 114 \zeta (ζ Greek char.), 502
\ (backslash), 167, 518 key, 9, 76 starts commands, 8, 14, 68 text symbol, 76–77, 81, 509 (space), 11, 63, 64 and \verb* command, 148 in arguments of commands, 85 in \bibitem labels, 266 in \cite commands, 264 in command names, 68 in tabular environments, 135 text symbol, 519 \ (space com.), 11, 66, 70, 102, 190, 520 ! (exclamation mark), 9, 62 float control, 260, 486 in \index commands, 453, 454, 457 ¡ (exclamation mark, Spanish), 81, 515, 518 \! (negthinspace), 182, 190, 190, 513, 520 \" (¨ dieresis/umlaut text accent), 9, 15, 79, 516, 548 " (double quote), 9, 75 in BibTEX database fields, 424, 445, 446 in \index commands, 457 key, 9, 63, 75 "ck (European character), 548 "s (eszett), 548 # in userdefined commands, 371, 378 key, 9, 63 \# (# octothorp), 77, 81, 518 $ as inline math delimiter, 17, 18, 33, 72, 74, 152 act as braces, 152 must be balanced, 153 in error messages, 20, 51, 147, 153, 154, 157, 233 key, 9, 63 \$ ($ dollar sign), 9, 77, 81, 518 $$ in error messages, 51 TEX displayed math delimiter, 152
618
%–@ %
as comment character, 6, 10, 19, 21, 61, 85–86, 141, 294, 458 in BibTEX databases, 86, 443 key, 9, 63 \% (% percent), 9, 61, 77, 81, 86, 518 in email addresses, 278 & as alignment point, 30, 31, 216, 218, 220, 221 as column separator, 25, 135, 220, 221, 231, 236 key, 9, 63 \& (& ampersand), 77, 81, 518 &&, alignment point for annotations, 32 ’ (right single quote), 9, 11, 62, 75, 519 for primes (0 ), 23, 159, 194 \’ ( ´ acute text accent), 79, 516 \( (start inline math mode), 18, 74 acts as special brace, 152, 153 must be balanced, 153 ( as math delimiter ( ( ), 167, 509 in index entries, 455 key, 8, 62 \) (end inline math mode), 18, 74 acts as special brace, 152, 153 must be balanced, 153 ) as math delimiter ( ) ), 167, 509 in index entries, 455 key, 8, 62 * (* asterisk), 77 interactive prompt, 113, 376–378, 534 key, 8, 63 ** prompt, 534 *form of commands, 46, 69, 98, 101, 103, 105, 129, 130, 148, 173, 195, 196, 242, 253, 322, 373, 382, 385, 439, 467, 473 of environments, 157, 201, 211, 213, 215, 231, 259 + key, 8, 63 + (plus), 22, 63, 157, 188, 191, 192, 213, 216–218 as binary operation, 506 + and − rule, 188, 191, 192, 217, 218 and subformulas, 216 in multline environments, 213 \, (thinspace), 25, 67, 75, 188, 190, 190, 513, 520 , (comma), 9, 62 and italic correction, 92 in \bibitem commands, 266
Index in BibTEX databases, 424, 425, 445 use in formulas, 155  (dash, hyphen, minus), 9, 22, 63, 75, 76, 157, 188, 191, 192, 196, 216, 217 as binary operation, 506 in \hyphenation commands, 82 \ (opt. hyphen), 13, 82, 82, 96  (– number ranges, en dash), 14, 75, 81, 518 in BibTEX databases, 427  (— em dash), 14, 75, 81, 518 . (period), 9, 62 and italic correction, 92 in BibTEX databases, 426 \. overdot text accent (˙), 79, 516 \/ (italic correction), 91–93, 520 and font commands, 92 / (slash) as divisor, 22, 63, 157 as math delimiter (/), 167, 509 key, 8 \: (medspace spacing com.), 190, 513, 520 : (colon), 9, 62 as binary relation, 63, 191, 503 \; (thickspace), 190, 513, 520 ; (semicolon), 9, 62 < (less than) as binary relation ( (tabbing com.), 141, 142 > (greater than) as binary relation (>), 503 key, 17, 63 text symbol, 81, 518 ? prompt, 20, 48, 491, 534 question mark, 9, 62 ¿ (Spanish question mark), 81, 515, 518 @ (@ at sign), 9, 63, 77 in BibTEX database (bib) files, 424 in \index commands, 455, 456 @expression, 136 @. (blank math symbol for comm. diagrams), 240, 241 \@. (intersentence space), 67, 520
Index
@–\\ @> (stretchable right arrow math sym.), 240 @AAA (stretchable up arrow math sym.), 241 @VVV (stretchable down arrow math sym.), 241 @\vert (stretchable double vertical line), 241 \[ (start displayed math mode), 18, 152, 155 acts as special brace, 152, 157 [ and optional arguments, 15, 27, 69, 70, 128, 132, 135 key, 8, 62 math delimiter ( [ ), 167, 509 with \item commands, 122 with \newcommand, 370 \] (end displayed math mode), 18, 152, 155 acts as special brace, 152, 157 ] and optional arguments, 15, 27, 69, 70, 128, 132, 135 key, 8, 62 math delimiter ( ] ), 167, 509 with \item commands, 122 with \newcommand, 370 { for required arguments, 8, 14, 19, 69, 70, 159, 366, 530 for scoping, 71–73 in BibTEX entries, 424, 427 key, 9, 63 must be balanced, 21, 72, 77 in \index commands, 462 with \def, 378 \{ as math delimiter ({ ), 167, 509 text brace ({ ), 77, 81, 518 } for required arguments, 8, 14, 19, 69, 70, 159, 366, 530 for scoping, 71–73 in BibTEX entries, 424, 427 key, 9, 63 must be balanced, 21, 72, 77 in \index commands, 462 with \def, 378 \} as math delimiter ( }), 167, 509
619 text brace ( }), 77, 81, 518 ˆ (caret) for superscripts, 23, 25, 158–159, 161, 182 key, 9, 63, 77 \^ circumflex text accent (ˆ), 79, 81, 516, 518 \ ( underscore), 9, 77, 81, 518 in email addresses, 278 (underscore) for subscripts, 23, 25, 158–159, 161, 178, 182 key, 9, 63 ‘ (left single quote), 9, 11, 62, 75, 519 \‘ ( ` grave text accent), 62, 79, 516 \ ( k math delimiter), 166, 167, 168, 185, 509  (  vertical line), 9, 17, 63, 77, 168, 189 as binary relation, 188, 189 as math delimiter, 167, 170, 189, 191, 509 in \index commands, 454, 457 in tabular environments, 135 ~ (tilde) key, 9, 11, 63 tie/unbreakable space, 11, 28, 63, 66, 75, 76, 99, 520 absorbs spaces, 76 in BibTEX databases, 427 with crossreferences, 28, 257 \~ ( ˜ tilde text accent), 77, 79, 516, 519 9pt (doc. class option), 297 10pt (doc. class option), 104, 297, 306 11pt (doc. class option), 297, 306 12pt (doc. class option), 93, 297, 306 \\ (new line), 15, 97, 98 and \kill commands, 142 breaking lines with, 25, 30, 31, 33, 44, 46, 98, 124, 135, 141, 142, 145, 176, 211, 213, 215, 216, 227, 231, 242, 273–278, 297, 304, 309, 358 in arguments of commands, 98 in environments, 98, 124, 145 in text, 98 optional argument of, 15, 98, 98, 242 \\* (new line), 98, 242