diff options
author | Karl Berry <karl@freefriends.org> | 2007-01-10 22:24:14 +0000 |
---|---|---|
committer | Karl Berry <karl@freefriends.org> | 2007-01-10 22:24:14 +0000 |
commit | 8a88f72a120a3d7f214cf0c24916e0c24da2a130 (patch) | |
tree | 47882be5f5f0fe051b894d8db556833a1f05f718 /Master/texmf-dist/doc/latex | |
parent | de0d34e2013ef10d645ab355c2a7f53510fc171c (diff) |
remove siam, selling not allowed
git-svn-id: svn://tug.org/texlive/trunk@3310 c570f23f-e606-0410-a88d-b1316a301751
Diffstat (limited to 'Master/texmf-dist/doc/latex')
-rw-r--r-- | Master/texmf-dist/doc/latex/siam/READ.ME | 38 | ||||
-rw-r--r-- | Master/texmf-dist/doc/latex/siam/docultex.tex | 592 | ||||
-rw-r--r-- | Master/texmf-dist/doc/latex/siam/lexample.tex | 658 |
3 files changed, 0 insertions, 1288 deletions
diff --git a/Master/texmf-dist/doc/latex/siam/READ.ME b/Master/texmf-dist/doc/latex/siam/READ.ME deleted file mode 100644 index d362d373370..00000000000 --- a/Master/texmf-dist/doc/latex/siam/READ.ME +++ /dev/null @@ -1,38 +0,0 @@ -READ.ME 12/11/92 - -The following files constitute the full distribution -of SIAM's LaTeX macro package. The files are: - -read.me -siamltex.sty -docultex.tex -siam10.sty -siam11.sty -numinsec.sty -subeqn.sty -siam.bst -lexample.tex - -docultex.tex is the documentation file and should be refered to -for information about using the macros. - -Lines 339, 340, 348, 349, and 351 of siamltex.sty need to be -commented out to work with the NFSS. - -The SIAM macro's home is NETLIB@research.att.com, accessible through email. - -Both the LaTeX and plain TeX macros can be accessed through NETLIB: -Send an e-mail message to netlib@research.att.com, the body of the -message should be: - - send siamltex.all from typesetting - -for the LaTeX version, and - - send siamptex.all from typesetting - -for the plain TeX version. - -Questions, problems, bugs, and bug fixes should be -forwarded to SIAM TeX Support, 3600 University City Science Center, -Philadelphia PA, 19104. E-mail to duggan@siam.org. diff --git a/Master/texmf-dist/doc/latex/siam/docultex.tex b/Master/texmf-dist/doc/latex/siam/docultex.tex deleted file mode 100644 index 811a058cc91..00000000000 --- a/Master/texmf-dist/doc/latex/siam/docultex.tex +++ /dev/null @@ -1,592 +0,0 @@ -%% -%% This is file `docultex.tex', -%% Documentation for siam macros for use with LaTeX 2e -%% -%% By Paul Duggan for the Society for Industrial and Applied -%% Mathematics. -%% -%% December 1, 1995 -%% -%% Version 1.0a -%% -%% You are not allowed to change this file. -%% -%% You are allowed to distribute this file under the condition that -%% it is distributed together with all of the files in the siam macro -%% distribution. These are: -%% -%% siamltex.cls (main LaTeX macro for SIAM) -%% siamltex.sty (includes siamltex.cls for compatibility mode) -%% siam10.clo (size option for 10pt papers) -%% subeqn.clo (allows equation numbners with lettered subelements) -%% siam.bst (bibliographic style file for BibTeX) -%% docultex.tex (this file) -%% lexample.tex (example file for latex macro) -%% -%% If you receive only some of these files from someone, complain! -%% -%% You are NOT ALLOWED to distribute this file alone. You are NOT -%% ALLOWED to take money for the distribution or use of either this -%% file or a changed version, except for a nominal charge for copying -%% etc. -%% \CharacterTable -%% {Upper-case \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 -%% Lower-case \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 -%% Digits \0\1\2\3\4\5\6\7\8\9 -%% Exclamation \! Double quote \" Hash (number) \# -%% Dollar \$ Percent \% Ampersand \& -%% Acute accent \' Left paren \( Right paren \) -%% Asterisk \* Plus \+ Comma \, -%% Minus \- Point \. Solidus \/ -%% Colon \: Semicolon \; Less than \< -%% Equals \= Greater than \> Question mark \? -%% Commercial at \@ Left bracket \[ Backslash \\ -%% Right bracket \] Circumflex \^ Underscore \_ -%% Grave accent \` Left brace \{ Vertical bar \| -%% Right brace \} Tilde \~} - -\documentclass[final]{siamltex} - -\title{USING SIAM'S \LaTeX\ MACROS\thanks{This work was -supported by the Society for Industrial and Applied -Mathematics}} - -\author{Paul Duggan\thanks{Society for Industrial and -Applied Mathematics, Philadelphia, Pennsylvania. -(duggan@siam. org). Questions, comments, or corrections -to this document may be directed to that email address.}} - - -\begin{document} -\maketitle - -\begin{abstract} -Documentation is given for use of the SIAM \LaTeX\ macros. These -macros are now compatible with \LaTeX$2_{\varepsilon}$. -Instructions and suggestions for compliance with SIAM style -standards are also included. Familiarity with standard \LaTeX\ commands -is assumed. -\end{abstract} - -\begin{keywords}\end{keywords} - -\begin{AMS}\end{AMS} - - -\pagestyle{myheadings} -\thispagestyle{plain} -\markboth{PAUL DUGGAN}{USING SIAM'S \LaTeX\ MACROS} - -\section{Introduction} - -This file is documentation for the SIAM \LaTeX\ macros, and -provides instruction for submission of your files. - -To accommodate authors who electronically typeset their manuscripts, -SIAM supports the use of \LaTeX. To ensure quality typesetting according -to SIAM style standards, SIAM provides a \LaTeX\ macro style file. -Using \LaTeX\ to format a manuscript should simplify the editorial process -and lessen the author's proofreading burden. However, -it is still necessary to proofread the galley proofs with care. - -Electronic files should not be submitted until the paper has been -accepted, and then not until requested to do so by someone in the SIAM -office. Once an article is slated for an issue, -someone from the SIAM office will contact the author about any or all -of the following: editorial and stylistic queries, -supplying the source files (and any supplementary macros) -for the properly formatted article, and handling figures. - -When submitting electronic files (electronic submissions) -(to {\tt tex@siam.org}) include the journal, issue, and author's -name in the subject line of the message. -Authors are responsible for ensuring that the paper generated -from the source files exactly matches the paper that -was accepted for publication by the review editor. If it does not, -information on how it differs should be indicated in the transmission -of the file. When submitting a file, please be sure to include any -additional macros (other than those provided by SIAM) that will be -needed to run the paper. - -SIAM uses MS-DOS-based computers for \LaTeX\ processing. Therefore -all filenames should be restricted to eight characters or less, -plus a three character extension. - -Once the files are corrected here at SIAM, we will mail the revised -proofs to be read against the original edited hardcopy -manuscript. We are not -set up to shuttle back and forth varying electronic versions of each -paper, so we must rely on hard copy of the galleys. The author's proofreading -is an important but easily overlooked step. Even if SIAM were not -to introduce a single editorial change into your manuscript, there -would still be a need to check, because electronic transmission -can introduce errors. - -The distribution contains the following items: {\tt -siamltex.cls}, the main macro package based on {\tt -article.cls}; {\tt siam10.clo}, for the ten-point size option;\linebreak -{\tt subeqn.clo}, a style option for equation numbering (see \S3 for -an explanation); and {\tt siam.bst}, - the style file for use with {\sc Bib}\TeX. Also included are this -file {\tt docultex.tex} and a sample file {\tt lexample.tex}. -The sample file represents a standard application of -the macros. The rest of this paper will highlight -some keys to effective macro use, as well as point out options and -special cases, and describe SIAM style standards to which -authors should conform. - - -\section{Headings} -The top matter of a journal paper falls into a standard -format. It begins of course with the \verb|\documentclass| command - -\begin{verbatim} - -\documentclass{siamltex} - -\end{verbatim} - -Other class options can be included -in the bracketed argument of the command, separated by commas. - - -The title and author parts are formatted using the -\verb|\title| and \verb|\author| commands as described in Lamport -\cite{Lamport}. The \verb|\date| -command is not used. \verb|\maketitle| produces the actual -output of the commands. - -The addresses and support acknowledgments are put into the -\verb|\author| commands via \verb|\thanks|. If support is -overall for the authors, the support acknowledgment should -be put in a \verb|\thanks| command in the \verb|\title|. -Specific support should go following the addresses of the -individual authors in the same \verb|\thanks| command. - -Sometimes authors have support or addresses in common which -necessitates having multiple \verb|\thanks| commands for -each author. Unfortunately \LaTeX\ does not normally allow this, -so a special procedure must be used. An example of this procedure -follows. Grant information can also be run into both authors' -footnotes. - -\begin{verbatim} - -\title{TITLE OF PAPER} - -\author{A.~U. Thorone\footnotemark[2]\ \footnotemark[5] -\and A.~U. Thortwo\footnotemark[3]\ \footnotemark[5] -\and A.~U. Thorthree\footnotemark[4]} - -\begin{document} -\maketitle - -\renewcommand{\thefootnote}{\fnsymbol{footnote}} - -\footnotetext[2]{Address of A.~U. Thorone} -\footnotetext[3]{Address of A.~U. Thortwo} -\footnotetext[4]{Address of A.~U. Thorthree} -\footnotetext[5]{Support in common for the first and second -authors.} - -\renewcommand{\thefootnote}{\arabic{footnote}} - -\end{verbatim} - -Notice that the footnote marks begin with {\tt [2]} -because the first mark (the asterisk) will be used in the -title for date-received information by SIAM, even if not -already used for support data. This is just one example; -other situations follow a similar pattern. - -Following the author and title is the abstract, key words -listing, and AMS subject classification number(s), -designated using the \verb|{abstract}|, \verb|{keywords}|, -and \verb|{AMS}| environments. If -there is only one AMS number, the commands -\verb|\begin{AM}| and \verb|\end{AM}| are used -instead of \verb|{AMS}|. This causes the heading to be -in the singular. Authors are responsible for providing AMS numbers. -They can be found in the Annual Index of Math Reviews, or -through {\tt e-Math} ({\tt telnet e-math.ams.com}; login -and password are both {\tt e-math}). - -Left and right running heads should be provided in the -following way. - -\begin{verbatim} - -\pagestyle{myheadings} -\thispagestyle{plain} -\markboth{A.~U. THORONE AND A.~U. THORTWO}{SHORTER PAPER -TITLE} - -\end{verbatim} - -\section{Equations and mathematics} - -One advantage of \LaTeX\ is that it can automatically number -equations and refer to these equation numbers in text. While plain \TeX's -method of equation numbering (explicit numbering using -\verb|\leqno|) works in the SIAM macro, it is not preferred -except in certain cases. SIAM style guidelines call for -aligned equations in many circumstances, and \LaTeX's -\verb|{eqnarray}| environment is not compatible with -\verb|\leqno| and \LaTeX\ is not compatible with the plain -\TeX\ command \verb|\eqalign| and \verb|\leqalignno|. Since -SIAM may have to alter or realign certain groups of -equations, it is necessary to use the \LaTeX\ system of -automatic numbering. - -Sometimes it is desirable to designate subequations of a larger -equation number. The subequations are designated with -(roman font) letters appended after the number. SIAM has -supplemented its macros with the {\tt subeqn.clo} option which -defines the environment \verb|{subequations}|. - -\begin{verbatim} - -\begin{subequations}\label{EKx} -\begin{equation} - y_k = B y_{k-1} + f, \qquad k=1,2,3,\ldots -\end{equation} -for any initial vector $ y_0$. Then -\begin{equation} - y_k\rightarrow u \mbox{\quad iff\quad} \rho( B)<1. -\end{equation} -\end{subequations} - -\end{verbatim} - -All equations within the \verb|{subequations}| environment -will keep the same overall number, but the letter -designation will increase. - -Clear equation formatting using \TeX\ can be challenging. Aside from -the regular \TeX\ documentation, authors will find Nicholas -J. Higham's book {\em Handbook of Writing for the Mathematical -Sciences\/} \cite{Higham} useful for guidelines and tips on -formatting with \TeX. The book covers many other topics related -to article writing as well. - -Authors commonly make mistakes by using - \verb|<|, \verb|>|, \verb|\mid|, and -\verb|\parallel| as delimiters, instead of -\verb|\langle|, \verb|\rangle|, \verb:|:, -and \verb:\|:. The incorrect symbols have particular -meanings distinct from the correct ones and should not be confused. - -\begin{table}[htbp] -\caption{Illustration of incorrect delimiter use.} -\begin{center}\footnotesize -\renewcommand{\arraystretch}{1.3} -\begin{tabular}{|ll|ll|}\hline -\multicolumn{2}{|c|}{{\bf Wrong}} & \multicolumn{2}{c|}{{\bf Right}}\\ \hline -\verb|<x, y>| & $<x, y>$ & \verb|\langle x, y\rangle| & $\langle x, y\rangle$\\ -\verb|5 < \mid A \mid| & $5 < \mid A \mid$ & \verb:5 < |A|: & $5 < |A|$\\ -\verb|6x = \parallel x|&&&\\ -\verb| - 1\parallel_{i}| & $6x = \parallel x - 1\parallel_{i}$ & - \verb:6x = \|x - 1\|_{i}: & $6x = \| x - 1\|_{i}$\\ \hline -\end{tabular} -\end{center} -\end{table} - -Another common author error is to put large (and even medium sized) -matrices in-line with the text, rather than displaying them. This -creates unattractive line spacing problems, and should be assiduously -avoided. Text-sized matrices (like $({a \atop b} {b \atop c})$) might -be used but anything much more complex than the example cited will -not be easy to read and should be displayed. - -More information on the formatting of equations and aligned -equations is found in Lamport \cite{Lamport}. Authors bear -primary responsibility for formatting their equations within -margins and in an aesthetically pleasing and informative manner. - -The SIAM macros include additional roman math words, or ``log-like" -functions, to those provided in standard \TeX. The following -commands are added: \verb:\const:, \verb:\diag:, \verb:\grad:, -\verb:\Range:, \verb:\rank:, and \verb:\supp:. -These commands produce the same word as the command name -in math mode, in upright type. - -\section{Special fonts} - -SIAM supports the use of the AMS-\TeX\ fonts (version 2.0 -and later). The package \verb|amsfonts| can be included with -the command\linebreak \verb|\usepackage{amsfonts}|. This package -is part of the AMS-\LaTeX distribution, available -from the AMS or from the Comprehensive TeX Archive -Network (anonymous ftp to ftp.shsu.edu). The blackboard bold font in this -font package can be used for designating number sets. -This is preferable to other methods of combining letters -(such as I and R for the real numbers) to produce pseudo-bold -letters but this is tolerable as well. Typographically speaking, -number sets may simply be designated using regular bold letters; -the blackboard bold typeface was designed to fulfil a desire -to simulate the limitations of a chalk board in printed type. - - -\subsection{Punctuation} -All standard punctuation and all numerals should be set in roman type -(upright) even within italic text. The only exceptions are periods and -commas. They may be set to match the surrounding text. - -References to sections should use the symbol \S, generated by -\verb|\S|. (If the reference begins a sentence, the term ``Section'' -should be spelled out in full.) Authors should not redefine -\verb|\S|, say, to be a calligraphic S, because \verb|\S| -must be reserved for use as the section symbol. - -Authors sometimes confuse the use of various types of dashes. -Hyphens (\verb|-|, -) are used for some compound words (many -such words should have no hyphen but must be run together, -like ``nonzero,'' or split apart, like ``well defined''). -Minus signs (\verb|$-$|, $-$) -should be used in math to represent subtraction or negative numbers. -En dashes (\verb|--|, --) are used for ranges (like 3--5, -June--August), or for joined names (like Runge--Kutta). Em dashes -(\verb|---|, ---) are used to set off a clause---such as this -one---from the rest of the sentence. - - -\subsection{Text formatting} - -SIAM style preferences do not make regular use of the \verb|{enumerate}| -and \verb|{itemize}| environments. Instead, -{\tt siamltex.cls} includes definitions of two alternate list -environments, \verb|{remunerate}| and \verb|{romannum}|. -Unlike the standard itemized lists, these environments do -not indent the secondary lines of text. The labels, whether -defaults or the optional user-defined, are always aligned -flush right. - -The \verb|{remunerate}| environment consecutively numbers -each item with an arabic numeral followed by a period. This -number is always upright, even in slanted -environments. (For those wondering at the unusual -naming of this environment, it comes from Seroul and Levy's -\cite{SerLev} definition of a similar macro for plain \TeX: -%{\tt \char"5C meti} -\verb|\meti| which is -% {\tt \char"5C item} -\protect\verb|\item| spelled backwards. Thus -%{\tt \{remunerate\}}, -\verb|{remunerate}| -a portion of -%{\tt \{enumerate\}} -\verb|{enumerate}| -spelled backwards.) - -The \verb|{romannum}| environment consecutively numbers -each item with a lower-case roman numeral enclosed in -parentheses. This number will always be upright within -slanted environments (as in theorems). - - -\section{Theorems and Lemmas} -Theorems, lemmas, corollaries, definitions, and propositions are covered -in the SIAM macros by the theorem-environments -\verb|{theorem}|, \verb|{lemma}|, \verb|{corollary}|, -\verb|{definition}| and \verb|{proposition}|. These are all -numbered in the same sequence and produce labels in small -caps with an italic body. Other environments may be specified by the -\verb|\newtheorem| command. SIAM's style is for Remarks and Examples -to appear with italic labels and an upright roman body. - -\begin{verbatim} - -\begin{theorem} -Sample theorem included for illustration. -Numbers and parentheses, like equation $(3.2)$, should be set -in roman type. Note that words (as opposed to ``log-like'' -functions) in displayed equations, such as -$$ x^2 = Y^2 \sin z^2 \mbox{ for all } x $$ -will appear in italic type in a theorem, though normally -they should appear in roman.\end{theorem} - -\end{verbatim} - -This sample produces Theorem 4.1 below. - -\begin{theorem} -Sample theorem included for illustration. -Numbers and parentheses, like equation $(3.2)$, should be set -in roman type. Note that words (as opposed to ``log-like'' -functions) in displayed equations, such as -$$ x^2 = Y^2 \sin z^2 \mbox{ for all } x $$ -will appear in italic type in a theorem, though normally -they should appear in roman. -\end{theorem} - - -Proofs are handled with the \verb|\begin{proof}| -\verb|\end{proof}| environment. A ``QED'' box \endproof\ is created -automatically by \verb|\end{proof}|, but this should be -preceded with a \verb|\qquad|. - -Named proofs, if used, must be done independently by the -authors. SIAM style specifies that proofs which end with -displayed equations should have the QED box two ems (\verb|\qquad|) -from the end of the equation on line with it horizontally. -Below is an example of how this can be done: - -\begin{verbatim} - -{\em Proof}. Proof of the previous theorem - . - . - . -thus, -$$ -a^2 + b^2 = c^2 \qquad\endproof -$$ - -\end{verbatim} - - -\section{Figures and tables} -Figures and tables sometimes require special consideration. -Tables in SIAM style are need to be set in eight point size -by using the \verb|\footnotesize| command inside the -\verb|\begin{table}| environment. Also, they should be designed -so that they do not extend beyond the text margins. - -SIAM style requires that no figures or tables appear in the -references section of the paper. \LaTeX\ is notorious for -making figure placement difficult, so it is important to -pay particular attention to figure placement near the -references in the text. All figures and tables should -be referred to in the text. - -SIAM supports the use of {\tt epsfig} for including {\sc PostScript} -figures. All {\sc Post\-Script} figures should be sent in separate -files. See the {\tt epsfig} documentation (available via -anonymous ftp from CTAN: ftp.shsu.edu) for more details on the use -of this style option. It is a good idea to submit high-quality -hardcopy of all {\sc Post\-Script} figures just in case there -is difficulty in the reproduction of the figure. Figures produced -by other non-\TeX\ methods should be included as high-quality -hardcopy when the manuscript is submitted. - -{\sc PostScript} figures that are sent should be generated with -sufficient line thickness. Some past figures authors have sent -had their line widths become very faint when SIAM set the papers -using a high-quality 1200dpi printer. - -Hardcopy for non-{\sc PostScript} figures should be included in -the submission of the hardcopy of the manuscript. Space -should be left in the \verb|{figure}| command for the -hardcopy to be inserted in production. - -\section{Bibliography and Bib\TeX} - -If using {\sc Bib}\TeX, authors need not submit the {\tt .bib} file for -their papers. Merely submit the completed {\tt .bbl} file, having used -{\tt siam.bst} as their bibliographic style file. {\tt siam.bst} -only works with Bib\TeX\ version 99i and later. The use of -Bib\TeX\ and the preparation of a {\tt .bib} file is -described in greater detail in \cite{Lamport}. - -If not using Bib\TeX, SIAM bibliographic references follow -the format of the following examples: - -\begin{verbatim} - -\bibitem{Ri} {\sc W. Riter}, -{\em Title of a paper appearing in a book}, in The Book -Title, E.~D. One, E.~D. Two, and A.~N. Othereditor, eds., -Publisher, Location, 1992, pp.~000--000. - -\bibitem{AuTh1} {\sc A.~U. Thorone}, {\em Title of paper -with lower case letters}, SIAM J. Abbrev. Correctly, 2 -(1992), pp.~000--000. - -\bibitem{A1A2} {\sc A.~U. Thorone and A.~U. Thortwo}, {\em -Title of paper appearing in book}, in Book Title: With All -Initial Caps, Publisher, Location, 1992. - -\bibitem{A1A22} \sameauthor, % generates the 3 em rule -{\em Title of Book{\rm :} Note Initial Caps and {\rm ROMAN -TYPE} for Punctuation and Acronyms}, Publisher, -Location, pp.~000--000, 1992. - -\bibitem{AuTh3} {\sc A.~U. Thorthree}, {\em Title of paper -that's not published yet}, SIAM. J. Abbrev. Correctly, to appear. - -\end{verbatim} - -Other types of references fall into the same general -pattern. See the sample file or any SIAM journal for other -examples. Authors must correctly format their bibliography to -be considered as having used the macros correctly. An incorrectly -formatted bibliography is not only time-consuming for SIAM to -process but it is possible that errors may be introduced into -it by keyboarders/copy editors. - -As an alternative to the above style of reference, an alphanumeric -code may be used in place of the number (e.g., [AUTh90]). The same -commands are used, but \verb|\bibitem| takes an optional argument -containing the desired alphanumeric code. - -Another alternative is no number, simply the authors' names and -the year of publication following in parentheses. The rest of the -format is identical. The macros do not support this alternative -directly, but modifications to the macro definition are possible -if this reference style is preferred. - - -\section{Conclusion} Many other style suggestions and tips -could be given to help authors but are beyond the scope of this -document. Simple mistakes can be avoided by increasing your familiarity -with how \LaTeX\ functions. The books referred to throughout this document -are also useful to the author who wants clear, beautiful typography -with minimal mistakes. - -\Appendix -\section{The use of appendices} -The \verb|\appendix| command may be used before the final sections -of a paper to designate them as appendices. Once \verb|\appendix| -is called, all subsequent sections will appear as - -\appendix -\section{Title of appendix} Each one will be sequentially lettered -instead of numbered. Theorem-like environments, subsections, -and equations will also have the section number changed to a letter. - -If there is only {\em one} appendix, however, the \verb|\Appendix| -(with a capital letter) should be used instead. This produces only -the word {\bf Appendix} in the section title, and does not add a letter. -Equation numbers, theorem numbers and subsections of the appendix -will have the letter ``A'' designating the section number. - -If you don't want to title your appendix, and just call it -{\bf Appendix A.} for example, use \verb|\appendix\section*{}| -and don't include anything in the title field. This works -opposite to the way \verb|\section*| usually works, by including the -section number, but not using a title. - - -Appendices should appear before the bibliography section, not after, -and any acknowledgments should be placed after the appendices and before -the bibliography. - -\begin{thebibliography}{1} -\bibitem{GoMiSa} {\sc M. Goossens, F. Mittelbach, and A. Samarin}, -{\em The} \LaTeX\ {\em Companion}, Addison-Wesley, Reading, MA, 1994. - -\bibitem{Higham} {\sc N.~J. Higham}, {\em Handbook of Writing for -the Mathematical Sciences}, Society for Industrial and Applied -Mathematics, Philadelphia, PA, 1993. - -\bibitem{Lamport} {\sc L. Lamport}, \LaTeX: {\em A Document -Preparation System}, Addison-Wesley, Reading, MA, 1986. - -\bibitem{SerLev} {\sc R. Seroul and S. Levy}, {\em A -Beginner's Book of} \TeX, Springer-Verlag, Berlin, New -York, 1991. -\end{thebibliography} - - -\end{document} -%% end of file `docultex.tex'
\ No newline at end of file diff --git a/Master/texmf-dist/doc/latex/siam/lexample.tex b/Master/texmf-dist/doc/latex/siam/lexample.tex deleted file mode 100644 index a8502d3788a..00000000000 --- a/Master/texmf-dist/doc/latex/siam/lexample.tex +++ /dev/null @@ -1,658 +0,0 @@ -%% -%% This is file `lexample.tex', -%% Sample file for siam macros for use with LaTeX 2e -%% -%% By Paul Duggan for the Society for Industrial and Applied -%% Mathematics. -%% -%% October 1, 1995 -%% -%% Version 1.0 -%% -%% You are not allowed to change this file. -%% -%% You are allowed to distribute this file under the condition that -%% it is distributed together with all of the files in the siam macro -%% distribution. These are: -%% -%% siamltex.cls (main LaTeX macro file for SIAM) -%% siamltex.sty (includes siamltex.cls for compatibility mode) -%% siam10.clo (size option for 10pt papers) -%% subeqn.clo (allows equation numbners with lettered subelements) -%% siam.bst (bibliographic style file for BibTeX) -%% docultex.tex (documentation file) -%% lexample.tex (this file) -%% -%% If you receive only some of these files from someone, complain! -%% -%% You are NOT ALLOWED to distribute this file alone. You are NOT -%% ALLOWED to take money for the distribution or use of either this -%% file or a changed version, except for a nominal charge for copying -%% etc. -%% \CharacterTable -%% {Upper-case \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 -%% Lower-case \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 -%% Digits \0\1\2\3\4\5\6\7\8\9 -%% Exclamation \! Double quote \" Hash (number) \# -%% Dollar \$ Percent \% Ampersand \& -%% Acute accent \' Left paren \( Right paren \) -%% Asterisk \* Plus \+ Comma \, -%% Minus \- Point \. Solidus \/ -%% Colon \: Semicolon \; Less than \< -%% Equals \= Greater than \> Question mark \? -%% Commercial at \@ Left bracket \[ Backslash \\ -%% Right bracket \] Circumflex \^ Underscore \_ -%% Grave accent \` Left brace \{ Vertical bar \| -%% Right brace \} Tilde \~} - - -\documentclass[final]{siamltex} - -% definitions used by included articles, reproduced here for -% educational benefit, and to minimize alterations needed to be made -% in developing this sample file. - -\newcommand{\pe}{\psi} -\def\d{\delta} -\def\ds{\displaystyle} -\def\e{{\epsilon}} -\def\eb{\bar{\eta}} -\def\enorm#1{\|#1\|_2} -\def\Fp{F^\prime} -\def\fishpack{{FISHPACK}} -\def\fortran{{FORTRAN}} -\def\gmres{{GMRES}} -\def\gmresm{{\rm GMRES($m$)}} -\def\Kc{{\cal K}} -\def\norm#1{\|#1\|} -\def\wb{{\bar w}} -\def\zb{{\bar z}} - -% some definitions of bold math italics to make typing easier. -% They are used in the corollary. - -\def\bfE{\mbox{\boldmath$E$}} -\def\bfG{\mbox{\boldmath$G$}} - -\title{Sample file for SIAM \LaTeX\ macro package\thanks{This - work was supported by the Society for Industrial and - Applied Mathematics, Philadelphia, Pennsylvania.}} - -% The thanks line in the title should be filled in if there is -% any support acknowledgement for the overall work to be included -% This \thanks is also used for the received by date info, but -% authors are not expected to provide this. - -\author{Paul Duggan\thanks{Composition Department, Society - for Industrial and Applied Mathematics, 3600 Univeristy - City Science Center, Philadelphia, Pennsylvania, - 19104-2688 ({\tt duggan@siam.org}).} - \and Various A.~U. Thors\thanks{Various Affiliations, - supported by various foundation grants.}} - -\begin{document} - -\maketitle - -\begin{abstract} -An example of SIAM \LaTeX\ macros is presented. Various -aspects of composing manuscripts for SIAM's journal series -are illustrated with actual examples from accepted -manuscripts. SIAM's stylistic standards are adhered to -throughout, and illustrated. -\end{abstract} - -\begin{keywords} -sign-nonsingular matrix, LU-factorization, indicator -polynomial -\end{keywords} - -\begin{AMS} -15A15, 15A09, 15A23 -\end{AMS} - -\pagestyle{myheadings} -\thispagestyle{plain} -\markboth{P. DUGGAN AND V. A. U. THORS}{SIAM MACRO EXAMPLES} - - -\section{Introduction and examples} -This paper presents a sample file for the use of SIAM's -\LaTeX\ macro package. It illustrates the features of the -macro package, using actual examples culled from various -papers published in SIAM's journals. It is to be expected -that this sample will provide examples of how to use the -macros to generate standard elements of journal papers, -e.g., theorems, definitions, or figures. This paper also -serves as an example of SIAM's stylistic preferences for -the formatting of such elements as bibliographic references, -displayed equations, and equation arrays, among others. -Some special circumstances are not dealt with in this -sample file; for such information one should see the -included documentation file. - -{\em Note:} This paper is not to be read in any form for content. -The conglomeration of equations, lemmas, and other text elements were -put together solely for typographic illustrative purposes and don't -make any sense as lemmas, equations, etc. - -\subsection{Sample text} -Let $S=[s_{ij}]$ ($1\leq i,j\leq n$) be a $(0,1,-1)$-matrix -of order $n$. Then $S$ is a {\em sign-nonsingular matrix} -(SNS-matrix) provided that each real matrix with the same -sign pattern as $S$ is nonsingular. There has been -considerable recent interest in constructing and -characterizing SNS-matrices \cite{bs}, \cite{klm}. There -has also been interest in strong forms of -sign-nonsingularity \cite{djd}. In this paper we give a new -generalization of SNS-matrices and investigate some of -their basic properties. - -Let $S=[s_{ij}]$ be a $(0,1,-1)$-matrix of order $n$ and -let $C=[c_{ij}]$ be a real matrix of order $n$. The pair -$(S,C)$ is called a {\em matrix pair of order} $n$. -Throughout, $X=[x_{ij}]$ denotes a matrix of order $n$ -whose entries are algebraically independent indeterminates -over the real field. Let $S\circ X$ denote the Hadamard -product (entrywise product) of $S$ and $X$. We say that the -pair $(S,C)$ is a {\em sign-nonsingular matrix pair of -order} $n$, abbreviated SNS-{\em matrix pair of order} $n$, -provided that the matrix \[A=S\circ X+C\] is nonsingular -for all positive real values of the $x_{ij}$. If $C=O$ -then the pair $(S,O)$ is a SNS-matrix pair if and only if -$S$ is a SNS-matrix. If $S=O$ then the pair $(O,C)$ is a -SNS-matrix pair if and only if $C$ is nonsingular. Thus -SNS-matrix pairs include both nonsingular matrices and -sign-nonsingular matrices as special cases. - -The pairs $(S,C)$ with -\[S=\left[\begin{array}{cc}1&0\\0&0\end{array}\right],\qquad -C=\left[\begin{array}{cc}1&1\\1&1\end{array}\right]\] and -\[S=\left[\begin{array}{ccc}1&1&0\\1&1&0\\0&0&0\end{array}\right],\qquad -C=\left[\begin{array}{ccc}0&0&1\\0&2&0\\ -3&0&0\end{array}\right]\] are examples of SNS-matrix pairs. - -\subsection{A remuneration list} -In this paper we consider the evaluation of integrals of the -following forms: -\begin{equation} -\int_a^b \left( \sum_i E_i B_{i,k,x}(t) \right) - \left( \sum_j F_j B_{j,l,y}(t) \right) dt,\label{problem} -\end{equation} -\begin{equation} -\int_a^b f(t) \left( \sum_i E_i B_{i,k,x}(t) \right) dt,\label{problem2} -\end{equation} -where $B_{i,k,x}$ is the $i$th B-spline of order $k$ defined over the -knots $x_i, x_{i+1}, \ldots, x_{i+k}$. -We will consider B-splines normalized so that their integral is one. -The splines may be of different orders and -defined on different knot sequences $x$ and $y$. -Often the limits of integration will be the entire real line, $-\infty$ -to $+\infty$. Note that (\ref{problem}) is a special case of (\ref{problem2}) -where $f(t)$ is a spline. - - -There are five different methods for calculating (\ref{problem}) -that will be considered: -\begin{remunerate} -\item Use Gauss quadrature on each interval. -\item Convert the integral to a linear combination of - integrals of products of B-splines and provide a recurrence for - integrating the product of a pair of B-splines. -\item Convert the sums of B-splines to piecewise - B\'{e}zier format and integrate segment - by segment using the properties of the Bernstein polynomials. -\item Express the product of a pair of B-splines as a linear combination - of B-splines. - Use this to reformulate the integrand as a linear combination - of B-splines, and integrate term by term. -\item Integrate by parts. -\end{remunerate} -Of these five, only methods 1 and 5 are suitable for calculating -(\ref{problem2}). The first four methods will be touched on and the -last will be discussed at length. - - -\subsection{Some displayed equations and \{{\tt eqnarray}\}s} - By introducing the product topology on $R^{m \times m} \times -R^{n \times n}$ with the induced inner product -\begin{equation} -\langle (A_{1},B_{1}), (A_{2},B_{2})\rangle := \langle A_{1},A_{2}\rangle -+ \langle B_{1},B_{2}\rangle,\label{eq2.10} -\end{equation} -we calculate the Fr\'{e}chet derivative of $F$ as follows: -\begin{eqnarray} - F'(U,V)(H,K) &=& \langle R(U,V),H\Sigma V^{T} + U\Sigma K^{T} - -P(H\Sigma V^{T} + U\Sigma K^{T})\rangle \nonumber \\ - &=& \langle R(U,V),H\Sigma V^{T} + U\Sigma K^{T}\rangle \label{eq2.11} \\ -&=& \langle R(U,V)V\Sigma^{T},H\rangle + \langle \Sigma^{T}U^{T}R(U,V),K^{T}\rangle. \nonumber -\end{eqnarray} -In the middle line of (\ref{eq2.11}) we have used the fact that the range of -$R$ is always perpendicular to the range of $P$. The gradient $\nabla F$ of -$F$, therefore, may be interpreted as the -pair of matrices: -\begin{equation} - \nabla F(U,V) = (R(U,V)V\Sigma^{T},R(U,V)^{T}U\Sigma ) \in -R^{m \times m} \times R^{n \times n}. \label{eq2.12} -\end{equation} -Because of the product topology, we know -\begin{equation} - {\cal T}_{(U,V)}({\cal O} (m) \times {\cal O} (n)) = -{\cal T}_{U}{\cal O} (m) \times {\cal T}_{V}{\cal O} (n), \label{eq2.13} -\end{equation} -where ${\cal T}_{(U,V)}({\cal O} (m) \times {\cal O} (n))$ stands for the -tangent space to the manifold ${\cal O} (m) \times {\cal O} (n)$ at $(U,V) -\in {\cal O} (m) \times {\cal O} (n)$ and so on. The projection of -$\nabla F(U,V)$ onto ${\cal T}_{(U,V)}({\cal O} (m) \times {\cal O} (n))$, -therefore, is the product of the projection of the first component of -$\nabla F(U,V)$ onto ${\cal T}_{U}{\cal O} (m)$ and the projection of the -second component of $\nabla F(U,V)$ onto ${\cal T}_{V}{\cal O} (n)$. -In particular, we claim that the -projection $ g(U,V)$ of the gradient $\nabla F(U,V)$ onto -${\cal T}_{(U,V)}({\cal O} (m) \times {\cal O} (n))$ is given by the pair of -matrices: -\begin{eqnarray} -g(U,V) = && \left( \frac{R(U,V)V\Sigma^{T}U^{T}-U\Sigma V^{T}R(U,V)^{T}}{2}U, -\right. \nonumber \\[-1.5ex] -\label{eq2.14}\\[-1.5ex] -&&\quad \left. \frac{R(U,V)^{T}U\Sigma V^{T}-V - \Sigma^{T}U^{T}R(U,V)}{2}V \right).\nonumber -\end{eqnarray} -Thus, the vector field -\begin{equation} -\frac{d(U,V)}{dt} = -g(U,V) \label{eq2.15} -\end{equation} -defines a steepest descent flow on the manifold ${\cal O} (m) \times -{\cal O} (n)$ for the objective function $F(U,V)$. - - -\section{Main results} - -Let $(S,C)$ be a matrix pair of order $n$. The determinant -\[\det (S\circ X+C)\] -is a polynomial in the indeterminates of $X$ of degree at -most $n$ over the real field. We call this polynomial the -{\em indicator polynomial} of the matrix pair $(S,C)$ -because of the following proposition. - -\begin{theorem} -\label{th:prop} -The matrix pair $(S,C)$ is a {\rm SNS}-matrix pair if and -only if all the nonzero coefficients in its indicator -polynomial have the same sign and there is at least one -nonzero coefficient. -\end{theorem} - -\begin{proof} -Assume that $(S,C)$ is a SNS-matrix pair. Clearly the -indicator polynomial has a nonzero coefficient. Consider a -monomial -\begin{equation} -\label{eq:mono} -b_{i_{1},\ldots,i_{k};j_{1},\ldots,j_{k}}x_{i_{1}j_{1}}\cdots -x_{i_{k}j_{k}} -\end{equation} -occurring in the indicator polynomial with a nonzero -coefficient. By taking the $x_{ij}$ that occur in -(\ref{eq:mono}) large and all others small, we see that any -monomial that occurs in the indicator polynomial with a -nonzero coefficient can be made to dominate all others. -Hence all the nonzero coefficients have the same sign. The -converse is im-\linebreak mediate. \qquad\end{proof} - - -For SNS-matrix pairs $(S,C)$ with $C=O$ the indicator -polynomial is a homogeneous polynomial of degree $n$. In -this case Theorem \ref{th:prop} is a standard fact about -SNS-matrices. - -\begin{lemma}[{\rm Stability}] -\label{stability} -Given $T>0$, suppose that $\| \epsilon (t) \|_{1,2} \leq h^{q-2}$ -for $0 \leq t \leq T$ and $q \geq 6$. -Then there exists a positive number $B$ that depends on -$T$ and the exact solution $\pe$ only such that for all $0 \leq t \leq T$, -\begin{equation} -\label{Gron} -\frac {d}{dt} \| \epsilon (t) \| _{1,2} \leq B - ( h^{q-3/2} + \| \epsilon (t) \|_{1,2})\;. -\end{equation} -The function $B(T)$ can be chosen to be nondecreasing in time. -\end{lemma} - - -\begin{theorem} -\label{th:gibson} -The maximum number of nonzero entries in a {\rm SNS}-matrix -$S$ of order $n$ equals \[\frac{n^{2}+3n-2}{2}\] with -equality if and only if there exist permutation matrices -such that $P|S|Q=T_{n}$ where -\begin{equation} -\label{eq:gibson} -T_{n}=\left[\begin{array}{cccccc} 1&1&\cdots&1&1&1\\ -1&1&\cdots&1&1&1\\ 0&1&\cdots&1&1&1\\ -\vdots&\vdots&\ddots&\vdots&\vdots&\vdots\\ -0&0&\cdots&1&1&1\\ 0&0&\cdots&0&1&1\end{array}\right]. -\end{equation} -\end{theorem} - -We note for later use that each submatrix of $T_{n}$ of -order $n-1$ has all 1s on its main diagonal. - -We now obtain a bound on the number of nonzero entries of -$S$ in a SNS-matrix pair $(S,C)$ in terms of the degree of -the indicator polynomial. We denote the strictly upper -triangular (0,1)-matrix of order $m$ with all 1s above the -main diagonal by $U_{m}$. The all 1s matrix of size $m$ by -$p$ is denoted by $J_{m,p}$. - - -\begin{proposition}[{\rm Convolution theorem}] -\label{pro:2.1} Let -\begin{eqnarray*} -a\ast u(t) = \int_0^t a(t- \tau) u(\tau) d\tau, \hspace{.2in} t \in -(0, \infty). -\end{eqnarray*} -Then -\begin{eqnarray*} -\widehat{a\ast u}(s) = \widehat{a}(s)\widehat{u}(s). -\end{eqnarray*} -\end{proposition} - -\begin{lemma} -\label{lem:3.1} -For $s_0 >0$, if -$$ -\int_0^{\infty} e^{-2s_0 t}v^{(1)}(t) v(t) dt \; \leq 0 \;, -$$ -then -\begin{eqnarray*} -\int_0^{\infty} e^{-2s_0 t} v^2(t) dt \; \leq \; \frac{1}{2s_0} v^2(0). -\end{eqnarray*} -\end{lemma} - -{\em Proof}. Applying integration by parts, we obtain -\begin{eqnarray*} -\int_0^{\infty} e^{-2s_0 t} [v^2(t)-v^2(0)] dt -&=&\lim_{t\rightarrow \infty}\left ( --\frac{1}{2s_0}e^{-2s_0 t}v^2(t) \right ) +\frac{1}{s_0} -\int_0^{\infty} e^{-2s_0 t}v^{(1)}(t)v(t)dt\\ -&\leq& \frac{1}{s_0} \int_0^{\infty} e^{-2s_0 t} v^{(1)}(t)v(t) dt \;\; -\leq \;\; 0. -\end{eqnarray*} -Thus -$$ -\int_0^{\infty} e^{-2s_0 t} v^2(t) dt \;\;\leq v^2(0) \int_0^{\infty} -\;\;e^{-2s_0 t} dt\;\;=\;\;\frac{1}{2s_0} v^2(0).\eqno\endproof -$$ - -\begin{corollary}\label{c4.1} -Let $ \bfE $ satisfy $(5)$--$(6)$ and -suppose $ \bfE^h $ satisfies $(7)$ and $(8)$ -with a general $ \bfG $. Let $ \bfG= \nabla \times {\bf \Phi} + \nabla p,$ -$p \in H_0^1 (\Omega) $. Suppose that $\nabla p$ and $ \nabla \times -{\bf \Phi} $ satisfy all the assumptions of Theorems $4.1$ and -$4.2$, respectively. In addition suppose all the regularity -assumptions of Theorems $4.1$--$4.2$ are satisfied. Then -for $ 0 \le t \le T $ and $ 0 < \epsilon \le \epsilon_0 $ there exists a -constant $ C = C(\epsilon, T) $ such that -$$ -\Vert (\bfE - \bfE^h)(t) \Vert_0 \le C h^{k+1- \epsilon}, -$$ -where $ C $ also depends on the constants given in Theorems -$4.1$ and $4.2$. -\end{corollary} - - -\begin{definition} -Let $S$ be an isolated invariant set with isolating neighborhood $N$. -An {\em index pair} for $S$ is a pair of compact sets $(N_{1},N_{0})$ -with $N_{0} \subset N_{1} \subset N$ such that: -\begin{romannum} -\item $cl(N_{1} \backslash N_{0})$ -is an isolating neighborhood for $S$. -\item $N_{i}$ is positively invariant relative to $N$ for $i=0,1$, -i.e., given -$x \in N_{i}$ and $x \cdot [0,t] \subset N$, then $x \cdot [0,t] \subset -N_{i}$. -\item $N_{0}$ is an exit set for $N_{1}$, i.e. if $x \in N_{1}$, -$x \cdot [0, \infty ) \not\subset N_{1}$, then there is a $T \geq 0$ such -that $x \cdot [0,T] \subset N_{1}$ and $x \cdot T \in N_{0}$. -\end{romannum} -\end{definition} - -\subsection{Numerical experiments} We conducted numerical experiments -in computing inexact Newton steps for discretizations of a -{\em modified Bratu problem}, given by -\begin{eqnarray} -{\ds \Delta w + c e^w + d{ {\partial w}\over{\partial x} } } -&=&{\ds f \quad {\rm in}\ D, }\nonumber\\[-1.5ex] -\label{bratu} \\[-1.5ex] -{\ds w }&=&{\ds 0 \quad {\rm on}\ \partial D , } \nonumber -\end{eqnarray} -where $c$ and $d$ are constants. The actual Bratu problem has $d=0$ and -$f \equiv0$. It provides a simplified model of nonlinear diffusion -phenomena, e.g., in combustion and semiconductors, and has been -considered by Glowinski, Keller, and Rheinhardt \cite{GloKR85}, -as well as by a number of other investigators; see \cite{GloKR85} -and the references therein. See also problem 3 by Glowinski and Keller -and problem 7 by Mittelmann in the collection of nonlinear model -problems assembled by Mor\'e \cite{More}. The modified problem -(\ref{bratu}) has been used as a test problem for inexact Newton -methods by Brown and Saad \cite{Brown-Saad1}. - -In our experiments, we took $D = [0,1]\times[0,1]$, $f \equiv0$, -$c=d=10$, and discretized (\ref{bratu}) using the usual second-order -centered differences over a $100\times100$ mesh of equally -spaced points in $D$. In \gmres($m$), we took $m=10$ and used fast -Poisson right preconditioning as in the experiments in \S2. The computing -environment was as described in \S2. All computing was done -in double precision. - - - -\begin{figure}[ht] -\vspace{2.5in} -\caption{{\rm Log}$_{10}$ of the residual norm versus the number of -{\rm GMRES$(m)$} iterations for the finite difference methods.} -\label{diff} -\end{figure} - - - -In the first set of experiments, we allowed each method to -run for $40$ {\gmresm} iterations, starting with zero as the initial -approximate solution, after which the limit of residual norm -reduction had been reached. The results are shown in Fig.~\ref{diff}. -In Fig.~\ref{diff}, the top curve was produced by method FD1. -The second curve from the top is actually a superposition of -the curves produced by methods EHA2 and FD2; the two curves are -visually indistinguishable. Similarly, the third curve from -the top is a superposition of the curves produced by methods EHA4 -and FD4, and the fourth curve from the top, which lies barely above -the bottom curve, is a superposition of the curves produced by -methods EHA6 and FD6. The bottom curve was produced by method A. - -In the second set of experiments, our purpose was to assess the -relative amount of computational work required by the methods -which use higher-order differencing to reach comparable levels -of residual norm reduction. We compared pairs of methods EHA2 -and FD2, EHA4 and FD4, and EHA6 and FD6 by observing in each of -20 trials the number of {\gmresm} iterations, number of $F$-evaluations, -and run time required by each method to reduce the residual norm -by a factor of $\e$, where for each pair of methods $\e$ was chosen -to be somewhat greater than the limiting ratio of final to -initial residual norms obtainable by the methods. In these trials, -the initial approximate solutions were obtained by generating random -components as in the similar experiments in \S2. We note that for every -method, the numbers of {\gmresm} iterations and $F$-evaluations required -before termination did not vary at all over the 20 trials. The {\gmresm} -iteration counts, numbers of $F$-evaluations, and means and standard -deviations of the run times are given in Table \ref{diffstats}. - -\begin{table} -\caption{Statistics over $20$ trials of {\rm GMRES$(m)$} iteration numbers, -$F$-evaluations, and run times required to reduce the residual norm by -a factor of $\e$. For each method, the number of {\rm GMRES$(m)$} iterations -and $F$-evaluations was the same in every trial.} - -\begin{center} \footnotesize -\begin{tabular}{|c|c|c|c|c|c|} \hline -&& Number of & Number of & Mean Run Time & Standard \\ -Method & $\e$ & Iterations & $F$-Evaluations& (Seconds) & Deviation \\ \hline -\lower.3ex\hbox{EHA2} & \lower.3ex\hbox{$10^{-10}$} & \lower.3ex\hbox{26} & -\lower.3ex\hbox{32} & \lower.3ex\hbox{47.12} & \lower.3ex\hbox{.1048} \\ -FD2 & $10^{-10}$ & 26 & 58 & 53.79 & .1829 \\ \hline -\lower.3ex\hbox{EHA4} & \lower.3ex\hbox{$10^{-12}$} & \lower.3ex\hbox{30} & -\lower.3ex\hbox{42} & \lower.3ex\hbox{56.76} & \lower.3ex\hbox{.1855} \\ -FD4 & $10^{-12}$ & 30 & 132 & 81.35 & .3730 \\ \hline -\lower.3ex\hbox{EHA6} & \lower.3ex\hbox{$10^{-12}$} & \lower.3ex\hbox{30} & -\lower.3ex\hbox{48} & \lower.3ex\hbox{58.56} & \lower.3ex\hbox{.1952} \\ -FD6 & $10^{-12}$ & 30 & 198 & 100.6 & .3278 \\ \hline -\end{tabular} -\end{center} -\label{diffstats} -\end{table} - -In our first set of experiments, we took $c=d=10$ and used right -preconditioning with a fast Poisson solver from {\fishpack} -\cite{Swarztrauber-Sweet}, which is very effective for these -fairly small values of $c$ and $d$. We first started each method -with zero as the initial approximate solution and allowed it -to run for 40 {\gmresm} iterations, after which the limit of residual -norm reduction had been reached. Figure \ref{pdep} shows plots -of the logarithm of the Euclidean norm of the residual versus -the number of {\gmresm} iterations for the three methods. We note -that in Fig.~\ref{pdep} and in all other figures below, the plotted -residual norms were not the values maintained by {\gmresm}, but rather -were computed as accurately as possible ``from scratch.'' That is, -at each {\gmresm} iteration, the current approximate solution was -formed and its product with the coefficient matrix was subtracted -from the right-hand side, all in double precision. -It was important to compute the residual norms in this way because -the values maintained by {\gmresm} become increasingly untrustworthy -as the limits of residual norm reduction are neared; see \cite{Walker88}. -It is seen in Fig.~\ref{pdep} that Algorithm EHA achieved -the same ultimate level of residual norm reduction as the FDP -method and required only a few more {\gmresm} iterations to do -so. - -\begin{figure}[t] -\vspace{3in} -\caption{{\rm Log}$_{10}$ of the residual norm versus the number of -{\rm GMRES}$(m)$ iterations for $c=d=10$ with fast Poisson -preconditioning. Solid curve: Algorithm {\rm EHA}; dotted -curve: {\rm FDP} method; dashed curve: {\rm FSP} method.} -\label{pdep} -\end{figure} - - -In our second set of experiments, we took $c=d=100$ and carried out -trials analogous to those in the first set above. No preconditioning -was used in these experiments, both because we wanted to compare -the methods without preconditioning and because the fast -Poisson preconditioning used in the first set of experiments is -not cost effective for these large values of $c$ and $d$. We first -allowed each method to run for 600 {\gmresm} iterations, -starting with zero as the initial approximate solution, after which -the limit of residual norm reduction had been reached. - - -\section*{Acknowledgments} -The author thanks the anonymous authors whose work largely -constitutes this sample file. He also thanks the INFO-TeX mailing -list for the valuable indirect assistance he received. - - -\begin{thebibliography}{10} -\bibitem{bs} {\sc R.~A. Brualdi and B.~L. Shader}, -{\em On sign-nonsingular matrices and the conversion of the -permanent into the determinant}, in Applied Geometry and -Discrete Mathematics, The Victor Klee Festschrift, P. -Gritzmann and B. Sturmfels, eds., American Mathematical -Society, Providence, RI, 1991, pp. 117--134. - -\bibitem{djd} {\sc J. Drew, C.~R. Johnson, and P. van den Driessche}, -{\em Strong forms of nonsingularity}, Linear Algebra Appl., -162 (1992), to appear. - -\bibitem{g} {\sc P.~M. Gibson}, {\em Conversion of the permanent into the -determinant}, Proc. Amer. Math. Soc., 27 (1971), -pp.~471--476. - -\bibitem{klm} -{\sc V.~Klee, R.~Ladner, and R.~Manber}, {\it -Signsolvability revisited}, Linear Algebra Appl., 59 -(1984), pp.~131--157. - -\bibitem{m} -{\sc K. Murota}, LU-{\em decomposition of a matrix with -entries of different kinds}, Linear Algebra Appl., 49 -(1983), pp.~275--283. - -\bibitem{Axelsson} -{\sc O.~Axelsson}, {\em Conjugate gradient type methods for unsymmetric and - inconsistent systems of linear equations}, Linear Algebra Appl., 29 (1980), - pp.~1--16. - -\bibitem{Brown-Saad1} -{\sc P.~N. Brown and Y.~Saad}, {\em Hybrid {K}rylov methods for nonlinear - systems of equations}, SIAM J. Sci. Statist. Comput., 11 (1990), - pp.~450--481. - -\bibitem{DES} -{\sc R.~S. Dembo, S.~C. Eisenstat, and T.~Steihaug}, {\em Inexact {N}ewton - methods}, SIAM J. Numer. Anal., 19 (1982), pp.~400--408. - -\bibitem{EES} -{\sc S.~C. Eisenstat, H.~C. Elman, and M.~H. Schultz}, {\em Variational - iterative methods for nonsymmetric systems of linear equations}, SIAM J. - Numer. Anal., 20 (1983), pp.~345--357. - -\bibitem{Elman} -{\sc H.~C. Elman}, {\em Iterative methods for large, sparse, nonsymmetric - systems of linear equations}, Ph.D. thesis, Department of Computer - Science, Yale University, New Haven, CT, 1982. - -\bibitem{GloKR85} -{\sc R.~Glowinski, H.~B. Keller, and L.~Rheinhart}, {\em Continuation-conjugate - gradient methods for the least-squares solution of nonlinear boundary value - problems}, SIAM J. Sci. Statist. Comput., 6 (1985), pp.~793--832. - -\bibitem{Golub-VanLoan} -{\sc G.~H. Golub and C.~F. Van~Loan}, {\em Matrix Computations}, - Second ed., The Johns Hopkins University Press, Baltimore, MD, 1989. - -\bibitem{More} -{\sc J.~J. Mor\'e}, {\em A collection of nonlinear model problems}, in - Computational Solutions of Nonlinear Systems of Equations, E.~L. Allgower and - K.~Georg, eds., Lectures in Applied Mathematics, Vol. 26, American - Mathematical Society, Providence, RI, 1990, pp.~723--762. - -\bibitem{Saad} -{\sc Y.~Saad}, {\em Krylov subspace methods for solving large unsymmetric - linear systems}, Math. Comp., 37 (1981), pp.~105--126. - -\bibitem{Saad-Schultz} -{\sc Y.~Saad and M.~H. Schultz}, {\em {\rm GMRES}: A generalized minimal - residual method for solving nonsymmetric linear systems}, SIAM J. Sci. Statist. - Comput., 7 (1986), pp.~856--869. - -\bibitem{Swarztrauber-Sweet} -{\sc P.~N. Swarztrauber and R.~A. Sweet}, {\em Efficient {\rm FORTRAN} - subprograms for the solution of elliptic partial differential equations}, ACM - Trans. Math. Software, 5 (1979), pp.~352--364. - -\bibitem{Walker88} -{\sc H.~F. Walker}, {\em Implementation of the {\rm GMRES} method using - {H}ouseholder transformations}, SIAM J. Sci. Statist. Comput., 9 (1988), - pp.~152--163. - -\bibitem{Walker89} -\sameauthor, {\em Implementations of - the {\rm GMRES} method}, Computer Phys. Comm., 53 (1989), pp.~311--320. - -\end{thebibliography} - -\end{document} - |