LATEX maths and graphics Tim Love May 26, 2004 This handout assumes that you have already read the Advanced LaTeX1 docum...

Author:
Love T.

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

LATEX maths and graphics Tim Love May 26, 2004 This handout assumes that you have already read the Advanced LaTeX1 document handout, so if you’re unsure about ‘environments’, read no further. Note that there’s an alternative way of producing maths in LATEX AMS-LATEX. See the online manual2 for details. If you want to more more about graphics, see Using Imported Graphics in LaTeX2e Documents3 by Keith Reckdahl. Comments and bug reports to Tim Love ([email protected]).

Contents 1

Maths 1.1 Environments . . . . . . . . . . . . . . . . . . 1.2 Special Characters . . . . . . . . . . . . . . . . 1.2.1 Greek . . . . . . . . . . . . . . . . . . . 1.2.2 Miscellaneous . . . . . . . . . . . . . . 1.2.3 Arrows . . . . . . . . . . . . . . . . . . 1.2.4 Calligraphic . . . . . . . . . . . . . . . 1.2.5 Character Modifiers . . . . . . . . . . 1.2.6 Common functions . . . . . . . . . . . 1.3 Subscripts and superscripts . . . . . . . . . . 1.4 Overlining, underlining and bold characters 1.5 Roots . . . . . . . . . . . . . . . . . . . . . . . 1.6 Fractions . . . . . . . . . . . . . . . . . . . . . 1.7 Delimiters . . . . . . . . . . . . . . . . . . . . 1.8 Numbering and labelling . . . . . . . . . . . 1.9 Matrices . . . . . . . . . . . . . . . . . . . . . 1.10 Macros . . . . . . . . . . . . . . . . . . . . . . 1.11 Packages . . . . . . . . . . . . . . . . . . . . . 1.12 Fine tuning . . . . . . . . . . . . . . . . . . . . 1.13 Maths and Postscript fonts . . . . . . . . . . . 1.14 Matlab and LaTeX . . . . . . . . . . . . . . . . 1.15 Examples . . . . . . . . . . . . . . . . . . . . . 1 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/latex

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

2 . 2 . 3 . 3 . 3 . 4 . 4 . 4 . 5 . 5 . 6 . 6 . 6 . 6 . 7 . 8 . 9 . 9 . 9 . 10 . 10 . 11

advanced/latex advanced.html

2 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/amslatex.dvi 3 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/epslatex.ps

c Copyright 2004 by T.P. Love. This document may be copied freely for the purposes of education and non-commercial research. Cambridge University Engineering Department, Cambridge CB2 1PZ, England. 1

2

Graphics 2.1 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 psfrag: adding maths to postscript files . . . 2.1.2 Postscript from PCs/Macs . . . . . . . . . . . . 2.2 Scaling, rotation, clipping, wrap-around and shadows 2.3 GIF and jpeg files . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

12 13 13 14 14 16

1 Maths There’s more to maths typesetting than meets the eye. Many conventions used in the typesetting of plain text are inappropriate to maths. LATEX goes a long way to help you along with the style. For example, in a LATEX maths environment, letters come out in italics, ‘-’ as ‘−’ (minus) instead of the usual ‘-’ (dash), ‘*’ becomes ∗, ’ becomes 0 and spacing is changed (less around ‘/’, more around ‘+’). Many of the usual LATEX constructions can still be used in maths environments but their effect may be slightly different; eg \textbf{ } only affects letters and numbers. ‘{’ and ‘}’ are still special characters; they’re used to group characters. As usual in LATEX you can override the defaults, but think before doing it: maths support in LATEX has been carefully thought out and is quite logical though the LATEX source text may not be very readable. It’s a good idea to write out the formulae on paper before you start LATEXing, and try not to overdo the use of the ‘\frac’ construction; use ‘/’ instead.

1.1

Environments

There are 2 environments to display one-line equations. equation:- Equations in this environment are numbered. \begin{equation} x + iy \end{equation} x + iy

(1)

displaymath:- These won’t be numbered. \[, \] can be used as abbreviations for \begin{displaymath} and \end{displaymath}. \begin{displaymath} x + iy \end{displaymath} x + iy Never leave a blank line before these equations; it starts a new paragraph and looks ugly. ’\displaystyle’ is the font type used to print maths in these display environments. Other relevant environments are:math:- For use in text. \( and \) can be used to delimit the environment, as can the TEX constructions $ and $ . For example, $x=yˆ2$ gives x = y 2 . eqnarray:- This is like a 3 column tabular environment. Each line by default is numbered. You can use the eqnarray* variant to suppress numbering altogether. 2

\begin{eqnarray} a1 & = & b1 + c1\nonumber\\ a2 & = & b2 - c2 \end{eqnarray} a1

= b1 + c1

a2

= b2 − c2

(2)

Maths in these 2 sorts of environments have different default sizes for some characters and other behavioural differences so that a line of maths won’t impinge on text lines below or above. If you want to put some non-maths text in amongst maths then enclose it in an \mbox{...}.

1.2

Special Characters

The amssymb package offers more symbols if the following aren’t enough. The symbols4 document (a postscript file) has a bigger list. 1.2.1 α ζ λ π φ ∆ Σ 1.2.2 ... .. . × ? · ∪ U F

∧ o 5 ⊕ ⊗ q ⊆ `

Greek \alpha \zeta \lambda \pi \phi \Delta \Sigma

β η µ ρ χ Θ Υ

\beta \eta \mu \rho \chi \Theta \Upsilon

γ θ ν σ ψ Λ Φ

\gamma \theta \nu \sigma \psi \Lambda \Phi

δ ι ξ τ ω Ξ Ψ

\delta \iota \xi \tau \omega \Xi \Psi

κ o υ Γ Π Ω

\epsilon \kappa o \upsilon \Gamma \Pi \Omega

Miscellaneous

\ldots

···

\cdots

\ddots \times \star \cdot \cup \biguplus \bigsqcup \wedge \wr \bigtriangledown \oplus \otimes \odot \amalg \preceq \subseteq \vdash \succeq

± ÷ ◦ ∩ S

\pm \div \circ \cap \bigcup \sqcap \vee \bigwedge \diamond \triangleleft \bigoplus \bigotimes \bigodot \leq \ll \sqsubseteq \geq \gg

u ∨ V

/L N J

≤ v ≥

.. . ∓ ∗ •T

] t W

\ 4 .

≺ ⊂ ∈ ⊃

4 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/symbols.ps

3

\vdots \mp \ast \bullet \bigcap \uplus \sqcup \bigvee \setminus \bigtriangleup \triangleright \ominus \oslash \bigcirc \prec \subset \in \succ \supset

⊇ a ' ∼ = ∝ | ^ h ¯ ` = ∇ ⊥ ∀ [ \ 4 `

1.2.3 ← ⇒ 7 → ) ⇐= ←→ ,→ ↑ ↓ % 1.2.4

\supseteq \dashv \simeq \cong \propto \mid \smile \hbar \ell \Im \nabla \bot \forall \flat \backslash \triangle \coprod Arrows

w ≡ 6= |= k _ ı ℘ 0√ k ∃ \ ∂ P R

\leftarrow \Rightarrow \mapsto \leftharpoondown \Longleftarrow \longleftrightarrow \hookrightarrow \uparrow \downarrow \nearrow \nwarrow

\sqsupseteq \equiv \asymp \neq \models \parallel \frown \imath \wp \prime \surd \| \exists \natural \partial \sum \int

⇐ ↔ ←* ) −→ ⇐⇒ * ⇑ ⇓ &

3 ∼ ≈ . = ⊥ ./ ℵ < > 6

¬ ] ∞ Q H

\ni \sim \approx \doteq \perp \bowtie \aleph \jmath \Re \empty \top \angle \neg \sharp \infty \prod \oint

\Leftarrow \leftrightarrow \hookleftarrow \rightleftharpoons \longrightarrow \Longleftrightarrow \rightharpoonup \Uparrow \Downarrow \searrow

→ ⇔ ( ←− =⇒ 7−→ +

\rightarrow \Leftrightarrow \leftharpoonup \longleftarrow \Longrightarrow \longmapsto \rightharpoondown

l .

\updownarrow \swarrow

Calligraphic

These characters are available if you use the \cal control sequence. ${\cal A B C D E F G H I J K L M N O P Q R S T U V W X Y Z}$ gives ABCDEFGHIJ KLMN OPQRST UVWX YZ 1.2.5

Character Modifiers \hat{e} \tilde{e} \check{e} \acute{e} \bar{e} \dot{e} \not e

eˆ e˜ eˇ e´ e¯ e˙ 6e

\widehat{easy} \widetilde{easy} \breve{e} \grave{e} \vec{e} \ddot{e}

easy d easy g e˘ e` ~e e¨

Note that the wide versions of hat and tilde cannot produce very wide alternatives. The ‘\not’ operator hasn’t properly cut the following letter. The Fine Tuning 4

section on page 9 describes how to adjust this. If you want to place one character above another, you can use \stackrel, which prints its first argument in small type immediately above the second $ a \stackrel{def}{=} b + c $ def

gives a = b + c See the Macros section for how to stack characters using atop. 1.2.6

Common functions

In a maths environment, LATEX assumes that variables will have single-character names. Function names require special treatment. The advantage of using the following control sequences for common functions is that the text will not be put in math italic and subscripts/superscripts will be made into limits where appropriate. \arccos \arcsin \arctan \arg \cos \cosh \cot \coth \csc \deg \det \dim \exp \gcd \hom \inf \ker \lg \lim \liminf \ln \log \max \min \Pr sec \sin \sinh \sup \tan \tanh

1.3

Subscripts and superscripts

These are introduced by the ‘ˆ’ and ‘_’ characters. Depending on the base character and the current style, the sub- or superscripts may go to the right of or directly above/below the main character. With letters it goes to the right. $F_2ˆ3$ produces ‘F23 ’. Note that the sub- and superscripts aren’t in line. To make them so, 3 you can add P an invisible character after the ‘F’. $F{}_2ˆ3$ produces F 2 . With the default behaviour is different in display and text styles.

$\sum_{i=0}ˆ2 $ P2 produces i=0 (text style) but \[\sum_{i=0}ˆ2 \]

produces (in display style) 2 X i=0

This default behaviour can be overridden, if you really need to. For example in text mode, $\sum\limits_{i=0}ˆ2$ produces

2 P

i=0

5

1.4

Overlining, underlining and bold characters

$\underline{one} \overline{two}$ produces onetwo. This is not a useful facility if it’s used more than once on a line. The lines are produced so that they don’t quite overlap the text; lines over or under different words won’t in general be at the same height. To be able to reproduce bold maths, it’s best to use the bm package. $E = \bm{mcˆ2}$ produces E = mc2 . Alternatively, you can use \mathbf{} to create bold characters - $\mathbf{F}_2ˆ3$ produces F32 . or you can use the following idea \usepackage{amsbsy} % This loads amstext too \begin{document} $\omega + \boldsymbol{\omega}$ % Use the following if whole expressions need to be in bold {\boldmath $\omega $} \end{document}

1.5

Roots

$\sqrt{4} + \sqrt[3]{x + y}$ √ √ gives 4 + 3 x + y.

1.6

Fractions

Three constructions for putting expressions above others are frac:- $\frac{1}{(x + 3)}$ produces

1 (x+3) .

choose:- ${n + 1 \choose 3}$ produces

n+1 3

atop:- ${x \atop y}$ produces xy .

.

These constructions can be used with ones described earlier. E.g., \[ \sum_{-1\le i \le 1 \atop 0 < j < \infty} f(i,j)\] gives

X

f (i, j)

−1≤i≤1 0<j0)\\ 0 & (x 0) (x < 0)

Over- and underbracing works too. $\overbrace{\alpha \ldots \omega}ˆ{\mbox{greek}} \underbrace{a \ldots z}_{\mbox{english}}$ greek z }| { produces α . . . ω a . . z} . The use of \mbox stops the text appearing in math italic. | .{z english

1.8

Numbering and labelling

Numbering happening automatically when you display equations. If you don’t want an equation numbered, use \nonumber beside the equation. Equation numbers appear to the right of the maths by default. To make them appear on the left use the leqno class option (i.e., use \documentclass[leqno,....]{....}). Use \label{} to label an equation (or figure, section etc) in order to reference from elsewhere.

7

\begin{equation} W_{\bf S}(t,\omega) = \int\limits_{-\infty}ˆ{\infty} { {\cal R}_{\bf S}(t,\tau) eˆ{-j\omega\tau} \,d \tau } \label{LABELLING} \end{equation} Z∞ WS (t, ω) = RS (t, τ )e−jωτ dτ

(3)

−∞

Now the following text refers back to equation \ref{LABELLING} refers back to equation 3 by number, and refers back to the equation on page \pageref{LABELLING} refers back to the equation on page 8. A file will have to be LATEX’ed twice before the references, both forwards and backwards, will be correctly produced.

1.9

Matrices

The array environment is like LATEX’s tabular environment except that each element is in math mode. The number and alignment of columns is controlled by the arguments - use l, c or r to represent each column with either left, center or right alignment. The default font style used is \textstyle but you can override this by changing the \displaystyle. \begin{math} \begin{array}{clrr} % a+b+c & uv & x-y & 27 \\ x+y & w & +z & 363 \end{array} \end{math} a + b + c uv x − y 27 x+y w +z 363 The rows are arranged so that their centres are aligned. You can align their tops or bottoms instead by using a further argument when you create the array. produces

\begin{array}{clrr}[t] would produce top-aligned lines, and ‘[b]’ would produce bottom-aligned ones. The Delimiters section of this document shows how to bracket matrices. TEX has a few maths facilities not mentioned in the LATEX book. The following TEX construction might be useful. \begin{math} \bordermatrix{&a_1&a_2&...&a_n\cr b_1 & 1.2 & 3.3 & 5.1 c_1 & 4.7 & 7.8 & 2.4 ... & ... & ... & ... z_1 & 8.0 & 9.9 & 0.9 \end{math} a a2 ... an 1 b1 1.2 3.3 5.1 2.8 c1 4.7 7.8 2.4 1.9 ... ... ... ... ... z1 8.0 9.9 0.9 9.99 8

& & & &

2.8 \cr 1.9 \cr ... \cr 9.99 \cr}

1.10

Macros

These aid readability, save on repetitive typing and offer ways of producing stylistic variations on standard LATEX formats. \def\bydefn{\stackrel{def}{=}} \def\convf{\hbox{\space \raise-2mm\hbox{$\textstyle \bigotimes \atop \scriptstyle \omega$} \space}} def

produce = and

N

when $\bydefn$ and $\convf$ are typed.

ω

1.11

Packages

The following packages may be of help • easybmat5 - easy block matrices • easyeqn6 - easy equations. • easymat7 - easy matrices • easytable8 - easy tables • easyvector9 - easy vectors • delarray10 - nested arrays • theorem11 - gives more choice in theorem layout • subeqnarray - Renumbering of sub-arrays in math-mode • subeqn - Different numbering sub-arrays

1.12

Fine tuning

It’s generally a good idea to keep punctuation outside math mode; LATEX’s normal handling of spacing around punctuation is suspended during maths. Sometimes you might want to adjust the spacing in a formula (e.g., you might want to add space before dx). Use these symbols :a\, b (a b) thin space a\> b (a b) medium space a\; b (a b) thick space a\! b (ab) negative thin space Long math expressions aren’t broken automatically unless you use the breqn 12 package, which is still a little experimental. In an eqnarray environment you may want to break a long line manually. You can do this by putting y & = & a + b \nonumber \\ & & + k 5 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easybmat.dvi 6 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easyeqn.dvi

7 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easymat.dvi

8 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easytable.dvi

9 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easyvector.dvi

10 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/delarray.dvi

11 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/theorem.dvi

12 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/breqndoc.dvi

9

to give y

=

a+b +k

(4)

but the spacing around the ‘+’ on the 2nd line is wrong because LATEX thinks it’s a unary operator. You can fool LATEX into treating it as a binary operator by inserting a hidden character. y & = & a + b \nonumber \\ & & \mbox{} + k gives y

=

a+b +k

(5)

You can use the \lefteqn construction to format long expressions so that continuation lines are differently indented. \begin{eqnarray} \lefteqn{x+ iy=}\\ & & a + b + c + d + e + f + g + h + i + j + k +\nonumber\\ & & l + m \nonumber \end{eqnarray} (6)

x + iy = a+b+c+d+e+f +g+h+i+j+k+ l+m

If you want more vertical spacing around a line you can create an invisible vertical ”struct” in LaTeX. \rule[-.3cm]{0cm}{1cm} creates a box of width 0, height 1cm which starts .3cm below the usual line base. By adjusting these values you should be able to create as much extra space below/above the maths as you A like. “ B and” is created by

$A \over B$

1.13

\rule[-.3cm]{0cm}{1cm}{and}

Maths and Postscript fonts

It’s easy to use a postscript font (like helvetica) for the text of a LATEX document. What’s harder is using the same font for maths. An easy, reasonable option is to use the mathptmx package to put the maths into the postscript Times and symbol fonts where possible. Alternatively, use • the mathpazo package (loads Palatino as the text font family and a mixture of the Pazo and CM fonts for math). • the mathpple package (loads Palatino as the text font family and a mixture of artificially obliqued Euler fonts and CM fonts for math). Commercial and free alternatives are under development.

1.14

Matlab and LaTeX

Matlab13 has some support for LaTeX production. For example 13 http://www-h.eng.cam.ac.uk/help/tpl/programs/matlab.html

10

latex(’(sin(x)+2*x+3*xˆ2)/(5*x+6*xˆ2)’,’math.tex’) puts the LaTeX representation of the expression into a file called math.tex. Type “help latex" inside matlab for details.

1.15

Examples

• \begin{equation} \hat{\theta}_{w_i} = \hat{\theta}(s(t,{\cal U}_{w_i})). \end{equation} gives ˆ Uwi )). θˆwi = θ(s(t,

(7)

• \begin{eqnarray} {\cal M}ˆ2(\hat{\theta},\theta) &=& E[(\hat{\theta} - \theta)ˆ2] \nonumber \\ {\cal M}ˆ2(\hat{\theta},\theta) &=& {\rm var}ˆ2(\hat{\theta}) + {\cal B}ˆ2(\hat{\theta}). \end{eqnarray} gives ˆ θ) M2 (θ, ˆ θ) M2 (θ,

= =

E[(θˆ − θ)2 ] ˆ + B 2 (θ). ˆ var2 (θ)

(8)

• \begin{equation} \hat{W}_{s}(t,\omega;\phi) \bydefn \int\limits_{-\infty}ˆ{\infty} {\hat{\cal R}_s(t,\tau;\psi) eˆ{-j\omega \tau} \, d \tau } \end{equation} gives ˆ s (t, ω; φ) def W =

Z∞

−∞

ˆ s (t, τ ; ψ)e−jωτ dτ R

• \begin{eqnarray} {\cal B}(t,\omega) & \approx & {1 \over 4\pi} {\cal D}_tˆ2 W_{\bf S}(t, \omega) {{{\scriptstyle \infty} \atop {\displaystyle \int \! \int \! }}\atop {\scriptstyle -\infty}} t_1ˆ2 \phi(t_1,\omega_1) \, dt_1 d\omega_1 \nonumber \\ && + {1 \over 4\pi} {\cal D}_\omegaˆ2 W_{\bf S}(t, \omega) {{{\scriptstyle \infty} \atop {\displaystyle \int \! \int \! }}\atop {\scriptstyle -\infty}} \omega_1ˆ2 \phi(t_1,\omega_1) \, dt_1 \, d\omega_1. 11

(9)

\label{F4} \end{eqnarray} gives

B(t, ω)

≈

1 2 D WS (t, ω) 4π t +

Z∞Z

t21 φ(t1 , ω1 ) dt1 dω1

−∞ Z∞Z

1 2 D WS (t, ω) 4π ω

ω12 φ(t1 , ω1 ) dt1 dω1 .

(10)

−∞

• \newsavebox{\DERIVBOXZLM} \savebox{\DERIVBOXZLM}[2.5em]{$\Longrightarrow\hspace{-1.5em} \raisebox{.2ex}{*} \hspace{-.7em}\raisebox{-.8ex}{\scriptsize lm}\hspace{.7em}$} \newcommand{\Deriveszlm}{\usebox{\DERIVBOXZLM}} \Deriveszlm gives * =⇒ lm

2 Graphics LATEX has a picture environment in which pictures can be drawn, but you’ll find graph paper handy. xfig can create code for the picture environment but the resulting graphics still suffer several limitations: only certain slopes and circles can be reproduced. The best method presently available on Unix is to use xfig to produce postscript files, which have no such limitations, but require a postscript printer or equivalent. Whatever graphics you want to add, you should use the figure environment so that LATEX can cope sensibly with situations where, for example, you attempt to insert near the bottom of a page a graphic that’s half a page high. The figure environment will float the graphic to the top or bottom of the page, or on the next page, with preferences that you can provide. h here t top of page b bottom of page p on a page with no text Putting ! as the first argument in the square brackets will encourage LATEX to do what you say, even if the result’s sub-optimal. See the online hints about floats in LaTeX14 for further details. \begin{figure}[htbp] \vspace{0.5in} \caption{0.5 inch of space} \end{figure} It’s possible to have more than one graphic in a figure. See the example later on. 14 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/float

12

hint.html

Figure 1: 0.5 inch of space

2.1

Postscript

pdflatex supports JPEG, PNG andPDF images - but not postscript. latex supports Postscript files as long as they have a proper bounding box comment; i.e. LATEX requires full Encapsulated Postscript15 as produced by (for example) xv and xfig on Unix. If the file hasn’t got a BoundingBox line near the top, you can use ps2epsi to generate one. Wherever the postscript file comes from, simply use \documentclass[dvips]{article} \usepackage{graphicx} then include the postscript file using the following commands \begin{figure}[htbp] \includegraphics{yourfile.ps} \end{figure} LATEX can cope with compressed postscript files too, but since latex can’t read the BoundingBox line from the compressed file, you need to provide it. If your compressed file’s called yourfile.ps.gz, copy the BoundingBox line into a file called yourfile.ps.bb. Then the following works \begin{figure}[htbp] \includegraphics{yourfile.ps.gz} \end{figure} Just about all of the following facilities use postscript. You’ll need to run latex to generate ‘foo.dvi’. This file can be viewed by the latest xdvi program, which can cope with embedded postscript. Run dvips -o foo.ps foo.dvi to convert the resulting DVI/postscript file to pure postscript. This will produce a file that can be previewed with ghostview or gs. On the teaching system this file can be printed out using plotview or lp. See the Creating and Printing graphics on PC, Mac, SUN and HP machines16 and Xfig17 handouts for more details. 2.1.1

psfrag: adding maths to postscript files

Many packages that produce postscript output don’t provide good maths facilities. It’s often easier to add the maths in later using the psfrag package. This lets you replace text in a postscript file (produced with xfig, matlab, etc) by a fragment of LATEX. For example, doing \usepackage{psfrag} ... \begin{figure} \psfrag{MATHS}{$xˆ2$} \includegraphics{foo.eps} \end{figure} 15 http://www-h.eng.cam.ac.uk/help/tpl/graphics/postscript.html

16 http://www-h.eng.cam.ac.uk/help/tpl/graphics/graphics2/graphics2.html 17 http://www-h.eng.cam.ac.uk/help/tpl/graphics/xfig/xfig.html

13

would display the file with MATHS replaced by x2 . See the online documentation18 for details. 2.1.2

Postscript from PCs/Macs

Modern applications should generate a conforming EPS file. Under windows when you’re printing to file, look at the PostScript properties (or Advanced options), and choose (depending on the driver you have) either ’Archive Format’, ’Encapsulated PS’, ’Optimize for Portability’ or ’Page Independence’. People seem to have more luck with the free Adobe postscript driver than with the Microsoft one. The resulting file should begin with %!PS. If it doesn’t it’s not postscript. Remove any characters that are before %!PS, and (to be on the safe side) remove anything after the final %%EOF line.Platform-specific considerations do crop up. The EPS generated on Macintoshes will use ASCII 13 line terminators, while Unix will use ASCII 10 (and DOS will use both). If this causes trouble, use emacs to convert, or try (in Unix) tr "\015" "\012" new.ps

2.2

Scaling, rotation, clipping, wrap-around and shadows

To scale, use some optional arguments \includegraphics[width=5cm,height=10cm]{yourfile.ps} would rescale the postscript so that it was 5cm wide and 10cm high. To make the picture 5cm wide and scale the height in proportion use \includegraphics[width=5cm]{yourfile.ps} To rotate anticlockwise by the specified number of degrees, use \includegraphics[angle=150]{yourfile.ps} These options can be combined - note that order matters. The following examples demonstrate how to combine these features and how to use the subfig package to have more than one graphic in a figure.

Figure 2: Tigers 18 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/psfrag.ps

14

\centering \begin{figure}[hbtp] \includegraphics[height=40mm]{/export/ghostfonts/tiger.eps} \includegraphics[angle=120, height=20mm]{/export/ghostfonts/tiger.eps} \caption{Tigers} \end{figure} % remember to do \usepackage{subfig} at the top of the document! \begin{figure}[hbtp] \centering \subfloat[Small]% \quad on the next line adds spacing {\includegraphics[height=30mm]{/export/ghostfonts/crest.eps}}\quad \subfloat[Medium] {\includegraphics[width=40mm]{CUni3.eps}}\quad \subfloat[Large] {\includegraphics[height=50mm]{/export/ghostfonts/crest.eps}} \caption{3 crests} \end{figure}

(a) Small

(b) Medium

(c) Large

Figure 3: 3 crests To clip the postscript image use the viewpoint argument. The following fragment would display only part of the image. The viewport coordinates are in the same units as the bounding box. \begin{figure}[htbp] \includegraphics[viewport=200 400 400 600,width=5cm,clip]% {CUniv3.eps} \end{figure} % Use the floatflt package \begin{floatingfigure}[l]{4cm} \includegraphics[width=2cm]{/export/ghostfonts/crest.eps} \caption{Using floatingfigure} \end{floatingfigure} The floatflt19 package lets you insert a graphic and have the text wrap around it. You can provide 2 arguments to the floatingfigure command: the first (l or r) selects whether you want the graphic to be on the left or right of the 19 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/floatflt.dvi

15

page. The 2nd argument gives the width of the graphic. Not all text will flow perfectly around (for example, verbatim text fails, as illustrated below) so check the final output carefully. Using the fancybox package gives you access to \shadowbox, \ovalbox, \Ovalbox and \doublebox commands, which can be used with text or with graphics. For example, \shadowbox{shadow package} shadow package produces and \ovalbox{\includegraphics[height=10mm]{CUni3.eps}} Figure 4: Using floating figure

produces . Unfortu nately, the fancybox package as supplied suppresses the table of contents. The locally produced contentsfancybox solves this, but may introduce graphics problems.

2.3

GIF and jpeg files

CUED users can include a file called keyboard.gif, for example, by doing gif2ps -b keyboard.gif (as long as you don’t change the GIF file you need only do this once) and then including the GIF file as you would a postscript one. For JPEG files run "jpeg2ps -h file.jpg > file.eps" then include the postscript file in the usual way. The resulting eps file will be little bigger than the original file. Alternatively, if you put the ’Bounding Box’ line from the eps file and put it in file.jpg.bb you can include JPEG files in the same way that you do postscript files.

16

Contents 1

Maths 1.1 Environments . . . . . . . . . . . . . . . . . . 1.2 Special Characters . . . . . . . . . . . . . . . . 1.2.1 Greek . . . . . . . . . . . . . . . . . . . 1.2.2 Miscellaneous . . . . . . . . . . . . . . 1.2.3 Arrows . . . . . . . . . . . . . . . . . . 1.2.4 Calligraphic . . . . . . . . . . . . . . . 1.2.5 Character Modifiers . . . . . . . . . . 1.2.6 Common functions . . . . . . . . . . . 1.3 Subscripts and superscripts . . . . . . . . . . 1.4 Overlining, underlining and bold characters 1.5 Roots . . . . . . . . . . . . . . . . . . . . . . . 1.6 Fractions . . . . . . . . . . . . . . . . . . . . . 1.7 Delimiters . . . . . . . . . . . . . . . . . . . . 1.8 Numbering and labelling . . . . . . . . . . . 1.9 Matrices . . . . . . . . . . . . . . . . . . . . . 1.10 Macros . . . . . . . . . . . . . . . . . . . . . . 1.11 Packages . . . . . . . . . . . . . . . . . . . . . 1.12 Fine tuning . . . . . . . . . . . . . . . . . . . . 1.13 Maths and Postscript fonts . . . . . . . . . . . 1.14 Matlab and LaTeX . . . . . . . . . . . . . . . . 1.15 Examples . . . . . . . . . . . . . . . . . . . . . 1 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/latex

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

2 . 2 . 3 . 3 . 3 . 4 . 4 . 4 . 5 . 5 . 6 . 6 . 6 . 6 . 7 . 8 . 9 . 9 . 9 . 10 . 10 . 11

advanced/latex advanced.html

2 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/amslatex.dvi 3 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/epslatex.ps

c Copyright 2004 by T.P. Love. This document may be copied freely for the purposes of education and non-commercial research. Cambridge University Engineering Department, Cambridge CB2 1PZ, England. 1

2

Graphics 2.1 Postscript . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 psfrag: adding maths to postscript files . . . 2.1.2 Postscript from PCs/Macs . . . . . . . . . . . . 2.2 Scaling, rotation, clipping, wrap-around and shadows 2.3 GIF and jpeg files . . . . . . . . . . . . . . . . . . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

12 13 13 14 14 16

1 Maths There’s more to maths typesetting than meets the eye. Many conventions used in the typesetting of plain text are inappropriate to maths. LATEX goes a long way to help you along with the style. For example, in a LATEX maths environment, letters come out in italics, ‘-’ as ‘−’ (minus) instead of the usual ‘-’ (dash), ‘*’ becomes ∗, ’ becomes 0 and spacing is changed (less around ‘/’, more around ‘+’). Many of the usual LATEX constructions can still be used in maths environments but their effect may be slightly different; eg \textbf{ } only affects letters and numbers. ‘{’ and ‘}’ are still special characters; they’re used to group characters. As usual in LATEX you can override the defaults, but think before doing it: maths support in LATEX has been carefully thought out and is quite logical though the LATEX source text may not be very readable. It’s a good idea to write out the formulae on paper before you start LATEXing, and try not to overdo the use of the ‘\frac’ construction; use ‘/’ instead.

1.1

Environments

There are 2 environments to display one-line equations. equation:- Equations in this environment are numbered. \begin{equation} x + iy \end{equation} x + iy

(1)

displaymath:- These won’t be numbered. \[, \] can be used as abbreviations for \begin{displaymath} and \end{displaymath}. \begin{displaymath} x + iy \end{displaymath} x + iy Never leave a blank line before these equations; it starts a new paragraph and looks ugly. ’\displaystyle’ is the font type used to print maths in these display environments. Other relevant environments are:math:- For use in text. \( and \) can be used to delimit the environment, as can the TEX constructions $ and $ . For example, $x=yˆ2$ gives x = y 2 . eqnarray:- This is like a 3 column tabular environment. Each line by default is numbered. You can use the eqnarray* variant to suppress numbering altogether. 2

\begin{eqnarray} a1 & = & b1 + c1\nonumber\\ a2 & = & b2 - c2 \end{eqnarray} a1

= b1 + c1

a2

= b2 − c2

(2)

Maths in these 2 sorts of environments have different default sizes for some characters and other behavioural differences so that a line of maths won’t impinge on text lines below or above. If you want to put some non-maths text in amongst maths then enclose it in an \mbox{...}.

1.2

Special Characters

The amssymb package offers more symbols if the following aren’t enough. The symbols4 document (a postscript file) has a bigger list. 1.2.1 α ζ λ π φ ∆ Σ 1.2.2 ... .. . × ? · ∪ U F

∧ o 5 ⊕ ⊗ q ⊆ `

Greek \alpha \zeta \lambda \pi \phi \Delta \Sigma

β η µ ρ χ Θ Υ

\beta \eta \mu \rho \chi \Theta \Upsilon

γ θ ν σ ψ Λ Φ

\gamma \theta \nu \sigma \psi \Lambda \Phi

δ ι ξ τ ω Ξ Ψ

\delta \iota \xi \tau \omega \Xi \Psi

κ o υ Γ Π Ω

\epsilon \kappa o \upsilon \Gamma \Pi \Omega

Miscellaneous

\ldots

···

\cdots

\ddots \times \star \cdot \cup \biguplus \bigsqcup \wedge \wr \bigtriangledown \oplus \otimes \odot \amalg \preceq \subseteq \vdash \succeq

± ÷ ◦ ∩ S

\pm \div \circ \cap \bigcup \sqcap \vee \bigwedge \diamond \triangleleft \bigoplus \bigotimes \bigodot \leq \ll \sqsubseteq \geq \gg

u ∨ V

/L N J

≤ v ≥

.. . ∓ ∗ •T

] t W

\ 4 .

≺ ⊂ ∈ ⊃

4 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/symbols.ps

3

\vdots \mp \ast \bullet \bigcap \uplus \sqcup \bigvee \setminus \bigtriangleup \triangleright \ominus \oslash \bigcirc \prec \subset \in \succ \supset

⊇ a ' ∼ = ∝ | ^ h ¯ ` = ∇ ⊥ ∀ [ \ 4 `

1.2.3 ← ⇒ 7 → ) ⇐= ←→ ,→ ↑ ↓ % 1.2.4

\supseteq \dashv \simeq \cong \propto \mid \smile \hbar \ell \Im \nabla \bot \forall \flat \backslash \triangle \coprod Arrows

w ≡ 6= |= k _ ı ℘ 0√ k ∃ \ ∂ P R

\leftarrow \Rightarrow \mapsto \leftharpoondown \Longleftarrow \longleftrightarrow \hookrightarrow \uparrow \downarrow \nearrow \nwarrow

\sqsupseteq \equiv \asymp \neq \models \parallel \frown \imath \wp \prime \surd \| \exists \natural \partial \sum \int

⇐ ↔ ←* ) −→ ⇐⇒ * ⇑ ⇓ &

3 ∼ ≈ . = ⊥ ./ ℵ < > 6

¬ ] ∞ Q H

\ni \sim \approx \doteq \perp \bowtie \aleph \jmath \Re \empty \top \angle \neg \sharp \infty \prod \oint

\Leftarrow \leftrightarrow \hookleftarrow \rightleftharpoons \longrightarrow \Longleftrightarrow \rightharpoonup \Uparrow \Downarrow \searrow

→ ⇔ ( ←− =⇒ 7−→ +

\rightarrow \Leftrightarrow \leftharpoonup \longleftarrow \Longrightarrow \longmapsto \rightharpoondown

l .

\updownarrow \swarrow

Calligraphic

These characters are available if you use the \cal control sequence. ${\cal A B C D E F G H I J K L M N O P Q R S T U V W X Y Z}$ gives ABCDEFGHIJ KLMN OPQRST UVWX YZ 1.2.5

Character Modifiers \hat{e} \tilde{e} \check{e} \acute{e} \bar{e} \dot{e} \not e

eˆ e˜ eˇ e´ e¯ e˙ 6e

\widehat{easy} \widetilde{easy} \breve{e} \grave{e} \vec{e} \ddot{e}

easy d easy g e˘ e` ~e e¨

Note that the wide versions of hat and tilde cannot produce very wide alternatives. The ‘\not’ operator hasn’t properly cut the following letter. The Fine Tuning 4

section on page 9 describes how to adjust this. If you want to place one character above another, you can use \stackrel, which prints its first argument in small type immediately above the second $ a \stackrel{def}{=} b + c $ def

gives a = b + c See the Macros section for how to stack characters using atop. 1.2.6

Common functions

In a maths environment, LATEX assumes that variables will have single-character names. Function names require special treatment. The advantage of using the following control sequences for common functions is that the text will not be put in math italic and subscripts/superscripts will be made into limits where appropriate. \arccos \arcsin \arctan \arg \cos \cosh \cot \coth \csc \deg \det \dim \exp \gcd \hom \inf \ker \lg \lim \liminf \ln \log \max \min \Pr sec \sin \sinh \sup \tan \tanh

1.3

Subscripts and superscripts

These are introduced by the ‘ˆ’ and ‘_’ characters. Depending on the base character and the current style, the sub- or superscripts may go to the right of or directly above/below the main character. With letters it goes to the right. $F_2ˆ3$ produces ‘F23 ’. Note that the sub- and superscripts aren’t in line. To make them so, 3 you can add P an invisible character after the ‘F’. $F{}_2ˆ3$ produces F 2 . With the default behaviour is different in display and text styles.

$\sum_{i=0}ˆ2 $ P2 produces i=0 (text style) but \[\sum_{i=0}ˆ2 \]

produces (in display style) 2 X i=0

This default behaviour can be overridden, if you really need to. For example in text mode, $\sum\limits_{i=0}ˆ2$ produces

2 P

i=0

5

1.4

Overlining, underlining and bold characters

$\underline{one} \overline{two}$ produces onetwo. This is not a useful facility if it’s used more than once on a line. The lines are produced so that they don’t quite overlap the text; lines over or under different words won’t in general be at the same height. To be able to reproduce bold maths, it’s best to use the bm package. $E = \bm{mcˆ2}$ produces E = mc2 . Alternatively, you can use \mathbf{} to create bold characters - $\mathbf{F}_2ˆ3$ produces F32 . or you can use the following idea \usepackage{amsbsy} % This loads amstext too \begin{document} $\omega + \boldsymbol{\omega}$ % Use the following if whole expressions need to be in bold {\boldmath $\omega $} \end{document}

1.5

Roots

$\sqrt{4} + \sqrt[3]{x + y}$ √ √ gives 4 + 3 x + y.

1.6

Fractions

Three constructions for putting expressions above others are frac:- $\frac{1}{(x + 3)}$ produces

1 (x+3) .

choose:- ${n + 1 \choose 3}$ produces

n+1 3

atop:- ${x \atop y}$ produces xy .

.

These constructions can be used with ones described earlier. E.g., \[ \sum_{-1\le i \le 1 \atop 0 < j < \infty} f(i,j)\] gives

X

f (i, j)

−1≤i≤1 0<j0)\\ 0 & (x 0) (x < 0)

Over- and underbracing works too. $\overbrace{\alpha \ldots \omega}ˆ{\mbox{greek}} \underbrace{a \ldots z}_{\mbox{english}}$ greek z }| { produces α . . . ω a . . z} . The use of \mbox stops the text appearing in math italic. | .{z english

1.8

Numbering and labelling

Numbering happening automatically when you display equations. If you don’t want an equation numbered, use \nonumber beside the equation. Equation numbers appear to the right of the maths by default. To make them appear on the left use the leqno class option (i.e., use \documentclass[leqno,....]{....}). Use \label{} to label an equation (or figure, section etc) in order to reference from elsewhere.

7

\begin{equation} W_{\bf S}(t,\omega) = \int\limits_{-\infty}ˆ{\infty} { {\cal R}_{\bf S}(t,\tau) eˆ{-j\omega\tau} \,d \tau } \label{LABELLING} \end{equation} Z∞ WS (t, ω) = RS (t, τ )e−jωτ dτ

(3)

−∞

Now the following text refers back to equation \ref{LABELLING} refers back to equation 3 by number, and refers back to the equation on page \pageref{LABELLING} refers back to the equation on page 8. A file will have to be LATEX’ed twice before the references, both forwards and backwards, will be correctly produced.

1.9

Matrices

The array environment is like LATEX’s tabular environment except that each element is in math mode. The number and alignment of columns is controlled by the arguments - use l, c or r to represent each column with either left, center or right alignment. The default font style used is \textstyle but you can override this by changing the \displaystyle. \begin{math} \begin{array}{clrr} % a+b+c & uv & x-y & 27 \\ x+y & w & +z & 363 \end{array} \end{math} a + b + c uv x − y 27 x+y w +z 363 The rows are arranged so that their centres are aligned. You can align their tops or bottoms instead by using a further argument when you create the array. produces

\begin{array}{clrr}[t] would produce top-aligned lines, and ‘[b]’ would produce bottom-aligned ones. The Delimiters section of this document shows how to bracket matrices. TEX has a few maths facilities not mentioned in the LATEX book. The following TEX construction might be useful. \begin{math} \bordermatrix{&a_1&a_2&...&a_n\cr b_1 & 1.2 & 3.3 & 5.1 c_1 & 4.7 & 7.8 & 2.4 ... & ... & ... & ... z_1 & 8.0 & 9.9 & 0.9 \end{math} a a2 ... an 1 b1 1.2 3.3 5.1 2.8 c1 4.7 7.8 2.4 1.9 ... ... ... ... ... z1 8.0 9.9 0.9 9.99 8

& & & &

2.8 \cr 1.9 \cr ... \cr 9.99 \cr}

1.10

Macros

These aid readability, save on repetitive typing and offer ways of producing stylistic variations on standard LATEX formats. \def\bydefn{\stackrel{def}{=}} \def\convf{\hbox{\space \raise-2mm\hbox{$\textstyle \bigotimes \atop \scriptstyle \omega$} \space}} def

produce = and

N

when $\bydefn$ and $\convf$ are typed.

ω

1.11

Packages

The following packages may be of help • easybmat5 - easy block matrices • easyeqn6 - easy equations. • easymat7 - easy matrices • easytable8 - easy tables • easyvector9 - easy vectors • delarray10 - nested arrays • theorem11 - gives more choice in theorem layout • subeqnarray - Renumbering of sub-arrays in math-mode • subeqn - Different numbering sub-arrays

1.12

Fine tuning

It’s generally a good idea to keep punctuation outside math mode; LATEX’s normal handling of spacing around punctuation is suspended during maths. Sometimes you might want to adjust the spacing in a formula (e.g., you might want to add space before dx). Use these symbols :a\, b (a b) thin space a\> b (a b) medium space a\; b (a b) thick space a\! b (ab) negative thin space Long math expressions aren’t broken automatically unless you use the breqn 12 package, which is still a little experimental. In an eqnarray environment you may want to break a long line manually. You can do this by putting y & = & a + b \nonumber \\ & & + k 5 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easybmat.dvi 6 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easyeqn.dvi

7 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easymat.dvi

8 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easytable.dvi

9 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/easyvector.dvi

10 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/delarray.dvi

11 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/theorem.dvi

12 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/breqndoc.dvi

9

to give y

=

a+b +k

(4)

but the spacing around the ‘+’ on the 2nd line is wrong because LATEX thinks it’s a unary operator. You can fool LATEX into treating it as a binary operator by inserting a hidden character. y & = & a + b \nonumber \\ & & \mbox{} + k gives y

=

a+b +k

(5)

You can use the \lefteqn construction to format long expressions so that continuation lines are differently indented. \begin{eqnarray} \lefteqn{x+ iy=}\\ & & a + b + c + d + e + f + g + h + i + j + k +\nonumber\\ & & l + m \nonumber \end{eqnarray} (6)

x + iy = a+b+c+d+e+f +g+h+i+j+k+ l+m

If you want more vertical spacing around a line you can create an invisible vertical ”struct” in LaTeX. \rule[-.3cm]{0cm}{1cm} creates a box of width 0, height 1cm which starts .3cm below the usual line base. By adjusting these values you should be able to create as much extra space below/above the maths as you A like. “ B and” is created by

$A \over B$

1.13

\rule[-.3cm]{0cm}{1cm}{and}

Maths and Postscript fonts

It’s easy to use a postscript font (like helvetica) for the text of a LATEX document. What’s harder is using the same font for maths. An easy, reasonable option is to use the mathptmx package to put the maths into the postscript Times and symbol fonts where possible. Alternatively, use • the mathpazo package (loads Palatino as the text font family and a mixture of the Pazo and CM fonts for math). • the mathpple package (loads Palatino as the text font family and a mixture of artificially obliqued Euler fonts and CM fonts for math). Commercial and free alternatives are under development.

1.14

Matlab and LaTeX

Matlab13 has some support for LaTeX production. For example 13 http://www-h.eng.cam.ac.uk/help/tpl/programs/matlab.html

10

latex(’(sin(x)+2*x+3*xˆ2)/(5*x+6*xˆ2)’,’math.tex’) puts the LaTeX representation of the expression into a file called math.tex. Type “help latex" inside matlab for details.

1.15

Examples

• \begin{equation} \hat{\theta}_{w_i} = \hat{\theta}(s(t,{\cal U}_{w_i})). \end{equation} gives ˆ Uwi )). θˆwi = θ(s(t,

(7)

• \begin{eqnarray} {\cal M}ˆ2(\hat{\theta},\theta) &=& E[(\hat{\theta} - \theta)ˆ2] \nonumber \\ {\cal M}ˆ2(\hat{\theta},\theta) &=& {\rm var}ˆ2(\hat{\theta}) + {\cal B}ˆ2(\hat{\theta}). \end{eqnarray} gives ˆ θ) M2 (θ, ˆ θ) M2 (θ,

= =

E[(θˆ − θ)2 ] ˆ + B 2 (θ). ˆ var2 (θ)

(8)

• \begin{equation} \hat{W}_{s}(t,\omega;\phi) \bydefn \int\limits_{-\infty}ˆ{\infty} {\hat{\cal R}_s(t,\tau;\psi) eˆ{-j\omega \tau} \, d \tau } \end{equation} gives ˆ s (t, ω; φ) def W =

Z∞

−∞

ˆ s (t, τ ; ψ)e−jωτ dτ R

• \begin{eqnarray} {\cal B}(t,\omega) & \approx & {1 \over 4\pi} {\cal D}_tˆ2 W_{\bf S}(t, \omega) {{{\scriptstyle \infty} \atop {\displaystyle \int \! \int \! }}\atop {\scriptstyle -\infty}} t_1ˆ2 \phi(t_1,\omega_1) \, dt_1 d\omega_1 \nonumber \\ && + {1 \over 4\pi} {\cal D}_\omegaˆ2 W_{\bf S}(t, \omega) {{{\scriptstyle \infty} \atop {\displaystyle \int \! \int \! }}\atop {\scriptstyle -\infty}} \omega_1ˆ2 \phi(t_1,\omega_1) \, dt_1 \, d\omega_1. 11

(9)

\label{F4} \end{eqnarray} gives

B(t, ω)

≈

1 2 D WS (t, ω) 4π t +

Z∞Z

t21 φ(t1 , ω1 ) dt1 dω1

−∞ Z∞Z

1 2 D WS (t, ω) 4π ω

ω12 φ(t1 , ω1 ) dt1 dω1 .

(10)

−∞

• \newsavebox{\DERIVBOXZLM} \savebox{\DERIVBOXZLM}[2.5em]{$\Longrightarrow\hspace{-1.5em} \raisebox{.2ex}{*} \hspace{-.7em}\raisebox{-.8ex}{\scriptsize lm}\hspace{.7em}$} \newcommand{\Deriveszlm}{\usebox{\DERIVBOXZLM}} \Deriveszlm gives * =⇒ lm

2 Graphics LATEX has a picture environment in which pictures can be drawn, but you’ll find graph paper handy. xfig can create code for the picture environment but the resulting graphics still suffer several limitations: only certain slopes and circles can be reproduced. The best method presently available on Unix is to use xfig to produce postscript files, which have no such limitations, but require a postscript printer or equivalent. Whatever graphics you want to add, you should use the figure environment so that LATEX can cope sensibly with situations where, for example, you attempt to insert near the bottom of a page a graphic that’s half a page high. The figure environment will float the graphic to the top or bottom of the page, or on the next page, with preferences that you can provide. h here t top of page b bottom of page p on a page with no text Putting ! as the first argument in the square brackets will encourage LATEX to do what you say, even if the result’s sub-optimal. See the online hints about floats in LaTeX14 for further details. \begin{figure}[htbp] \vspace{0.5in} \caption{0.5 inch of space} \end{figure} It’s possible to have more than one graphic in a figure. See the example later on. 14 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/float

12

hint.html

Figure 1: 0.5 inch of space

2.1

Postscript

pdflatex supports JPEG, PNG andPDF images - but not postscript. latex supports Postscript files as long as they have a proper bounding box comment; i.e. LATEX requires full Encapsulated Postscript15 as produced by (for example) xv and xfig on Unix. If the file hasn’t got a BoundingBox line near the top, you can use ps2epsi to generate one. Wherever the postscript file comes from, simply use \documentclass[dvips]{article} \usepackage{graphicx} then include the postscript file using the following commands \begin{figure}[htbp] \includegraphics{yourfile.ps} \end{figure} LATEX can cope with compressed postscript files too, but since latex can’t read the BoundingBox line from the compressed file, you need to provide it. If your compressed file’s called yourfile.ps.gz, copy the BoundingBox line into a file called yourfile.ps.bb. Then the following works \begin{figure}[htbp] \includegraphics{yourfile.ps.gz} \end{figure} Just about all of the following facilities use postscript. You’ll need to run latex to generate ‘foo.dvi’. This file can be viewed by the latest xdvi program, which can cope with embedded postscript. Run dvips -o foo.ps foo.dvi to convert the resulting DVI/postscript file to pure postscript. This will produce a file that can be previewed with ghostview or gs. On the teaching system this file can be printed out using plotview or lp. See the Creating and Printing graphics on PC, Mac, SUN and HP machines16 and Xfig17 handouts for more details. 2.1.1

psfrag: adding maths to postscript files

Many packages that produce postscript output don’t provide good maths facilities. It’s often easier to add the maths in later using the psfrag package. This lets you replace text in a postscript file (produced with xfig, matlab, etc) by a fragment of LATEX. For example, doing \usepackage{psfrag} ... \begin{figure} \psfrag{MATHS}{$xˆ2$} \includegraphics{foo.eps} \end{figure} 15 http://www-h.eng.cam.ac.uk/help/tpl/graphics/postscript.html

16 http://www-h.eng.cam.ac.uk/help/tpl/graphics/graphics2/graphics2.html 17 http://www-h.eng.cam.ac.uk/help/tpl/graphics/xfig/xfig.html

13

would display the file with MATHS replaced by x2 . See the online documentation18 for details. 2.1.2

Postscript from PCs/Macs

Modern applications should generate a conforming EPS file. Under windows when you’re printing to file, look at the PostScript properties (or Advanced options), and choose (depending on the driver you have) either ’Archive Format’, ’Encapsulated PS’, ’Optimize for Portability’ or ’Page Independence’. People seem to have more luck with the free Adobe postscript driver than with the Microsoft one. The resulting file should begin with %!PS. If it doesn’t it’s not postscript. Remove any characters that are before %!PS, and (to be on the safe side) remove anything after the final %%EOF line.Platform-specific considerations do crop up. The EPS generated on Macintoshes will use ASCII 13 line terminators, while Unix will use ASCII 10 (and DOS will use both). If this causes trouble, use emacs to convert, or try (in Unix) tr "\015" "\012" new.ps

2.2

Scaling, rotation, clipping, wrap-around and shadows

To scale, use some optional arguments \includegraphics[width=5cm,height=10cm]{yourfile.ps} would rescale the postscript so that it was 5cm wide and 10cm high. To make the picture 5cm wide and scale the height in proportion use \includegraphics[width=5cm]{yourfile.ps} To rotate anticlockwise by the specified number of degrees, use \includegraphics[angle=150]{yourfile.ps} These options can be combined - note that order matters. The following examples demonstrate how to combine these features and how to use the subfig package to have more than one graphic in a figure.

Figure 2: Tigers 18 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/psfrag.ps

14

\centering \begin{figure}[hbtp] \includegraphics[height=40mm]{/export/ghostfonts/tiger.eps} \includegraphics[angle=120, height=20mm]{/export/ghostfonts/tiger.eps} \caption{Tigers} \end{figure} % remember to do \usepackage{subfig} at the top of the document! \begin{figure}[hbtp] \centering \subfloat[Small]% \quad on the next line adds spacing {\includegraphics[height=30mm]{/export/ghostfonts/crest.eps}}\quad \subfloat[Medium] {\includegraphics[width=40mm]{CUni3.eps}}\quad \subfloat[Large] {\includegraphics[height=50mm]{/export/ghostfonts/crest.eps}} \caption{3 crests} \end{figure}

(a) Small

(b) Medium

(c) Large

Figure 3: 3 crests To clip the postscript image use the viewpoint argument. The following fragment would display only part of the image. The viewport coordinates are in the same units as the bounding box. \begin{figure}[htbp] \includegraphics[viewport=200 400 400 600,width=5cm,clip]% {CUniv3.eps} \end{figure} % Use the floatflt package \begin{floatingfigure}[l]{4cm} \includegraphics[width=2cm]{/export/ghostfonts/crest.eps} \caption{Using floatingfigure} \end{floatingfigure} The floatflt19 package lets you insert a graphic and have the text wrap around it. You can provide 2 arguments to the floatingfigure command: the first (l or r) selects whether you want the graphic to be on the left or right of the 19 http://www-h.eng.cam.ac.uk/help/tpl/textprocessing/floatflt.dvi

15

page. The 2nd argument gives the width of the graphic. Not all text will flow perfectly around (for example, verbatim text fails, as illustrated below) so check the final output carefully. Using the fancybox package gives you access to \shadowbox, \ovalbox, \Ovalbox and \doublebox commands, which can be used with text or with graphics. For example, \shadowbox{shadow package} shadow package produces and \ovalbox{\includegraphics[height=10mm]{CUni3.eps}} Figure 4: Using floating figure

produces . Unfortu nately, the fancybox package as supplied suppresses the table of contents. The locally produced contentsfancybox solves this, but may introduce graphics problems.

2.3

GIF and jpeg files

CUED users can include a file called keyboard.gif, for example, by doing gif2ps -b keyboard.gif (as long as you don’t change the GIF file you need only do this once) and then including the GIF file as you would a postscript one. For JPEG files run "jpeg2ps -h file.jpg > file.eps" then include the postscript file in the usual way. The resulting eps file will be little bigger than the original file. Alternatively, if you put the ’Bounding Box’ line from the eps file and put it in file.jpg.bb you can include JPEG files in the same way that you do postscript files.

16

Our partners will collect data and use cookies for ad personalization and measurement. Learn how we and our ad partner Google, collect and use data. Agree & close