MATLAB FAQ (Apr 15,2003)

Genial

MATLAB FAQ (Apr 15,2003)
http://www.mit.edu/%7Epwb/cssm/matlab-faq.txt
                                   MATLAB FAQ
  
Frequently asked questions in comp.soft-sys.matlab
  
     Last modified: <Tue Apr 15 13:00:54 2003 by pwb>
  
     Recent Changes:
  
    (+) New Question
    (%) Modified Question
      * (+) Q3.12: How can I process a sequence of files?
      _________________________________________________________________
  
     Table of Contents
  
      * Introduction
      * 1 Introduction, Policy, Credits
           + Q1.1: What is MATLAB?
           + Q1.2: What is this newsgroup about?
           + Q1.3: Where can I find an archive of comp.soft-sys.matlab?
           + Q1.4: What other software packages exist to do similar work?
           + Q1.5: What is the FAQ editorial policy?
           + Q1.6: Who maintains this FAQ?
           + Q1.7: Contributors
      * 2 Basics
           + Q2.1: Why does MATLAB only calculate to 4 significant digits?
           + Q2.2:  Why  does  the  transpose  operator  take  the complex
             conjugate?
           + Q2.3: How can I set all NaN's to 0?
           + Q2.4:  How  can  I  make MATLAB use the full window width for
             displaying matrices?
           + Q2.5: How do I comment out a large block of code?
           + Q2.6: How do I save default settings across sessions?
           + Q2.7: How can I find local maxima in a vector array?
           + Q2.8: Why is 6*i not complex in my program?
           + Q2.9: How can I modify the MATLAB path?
      * 3 Programming
           + Q3.1: Can I read a text file with a certain format?
           + Q3.2: Why are global variables bad?
           + Q3.3: What about this logical array business?
           + Q3.4: Huge memory waste using array of structs?
           + Q3.5: Why is my MEX file crashing?
           + Q3.6: How can I create variables A1, A2,...,A10 in a loop?
           + Q3.7: Do boolean operators short-circuit?
           + Q3.8: How do I fix "Out of Memory" problems?
           + Q3.9: How do I dynamically generate a filename for SAVE?
           + Q3.10:  What's the difference between M-files, Pcode, and MEX
             files?
           + Q3.11: Can MATLAB pass by reference?
           + Q3.12: How can I process a sequence of files?
      * 4 Graphics
           + Q4.1: How do I adjust the fontsize of a ticklabel?
           + Q4.2: Can I create a pi/sigma/superscript in my ticklabels?
           + Q4.3: Can I open multiple files using uigetfile?
           + Q4.4:   I   want  to  use  a  scrollbar  to  scroll  my  edit
             boxes/buttons
           + Q4.5: How can I rotate ticklabels?
           + Q4.6: How can I display data in a grid like Excel?
           + Q4.7: How can I write a title or text with multiple lines?
           + Q4.8:  How  can  I  use  a  common  color scale among several
             images?
           + Q4.9: How can I set default handle graphics properties?
           + Q4.10: How can I modify the default background color used for
             histograms?
           + Q4.11: How can I draw more than two lines with plotyy?
           + Q4.12:  Is there a command available for plotting an arrow on
             a graph?
           + Q4.13: Why does movie(M,1) display the movie M twice?
           + Q4.14: How can I set the focus in my GUI?
           + Q4.15: How can I add text labels to data points?
      * 5 Math/Algorithms
           + Q5.1: Why is 0.3-0.2-0.1 not equal to zero (or similar)?
           + Q5.2: How does the backslash operator work? What does it do?
           + Q5.3: How does one compute a factorial?
           + Q5.4: How can one accurately compute ratios of factorials?
           + Q5.5: How can I fit a circle to a set of XY data?
           + Q5.6: Why does MATLAB return an complex number for (-8)^(1/3)
           + Q5.7: Can I compute the DFT of an irregularly-sampled signal?
           + Q5.8:   How  can  I  efficiently  do  row-wise  vector-matrix
             multiplication?
      * 6 Installation, Crashes, Platform-Specific Issues
           + Q6.1: MATLAB 5.3 won't run on my linux system.
           + Q6.2:  Why  do  I  get a segfault after compiling correct MEX
             files under linux?
           + Q6.3: Why does MATLAB 6.0 crash on my Pentium 4?
           + Q6.4: How can I make MATLAB 6.0 (R12) open without the GUI?
           + Q6.5: Why does MATLAB 6.0 crash under Redhat 7.0?
           + Q6.6: Why can't I type into the command window?
           + Q6.7: What happened to MATLAB for the Macintosh?
           + Q6.8: How do I release a MATLAB license?
           + Q6.9: Is MATLAB optimized for the Pentium IV?
           + Q6.10: Why doesn't Ctrl-C interrupt execution of my m-file?
      * 7 Toolboxes, Software Archives, Books, and Other Resources
           + Q7.1: Where can I find archives of MATLAB software?
           + Q7.2: Where can I find a package to perform a specific task?
           + Q7.3: What textbook would you recommend for learning Matlab?
           + Q7.4:  Where  can  I  find  MATLAB  Style  Guides  or  Coding
             Standards?
      * 8 Simulink
           + Q8.1:  How  do  I  insert  a Simulink model into an MS Office
             Document?
      * 9 Miscellaneous
           + Q9.1: Why does MATLAB 6.0 (R12) take so much memory?
           + Q9.2: Can I use MATLAB to interface to an RS232 serial port?
           + Q9.3:  I've  seen  system_dependent  on the newsgroup before.
             What does it do?
           + Q9.4: Why does bench run so much slower in MATLAB 6.x than in
             5.x?
           + Q9.5: What is vectorization?
           + Q9.6: What is Acklamization?
           + Q9.7: Why does MATLAB needlessly access the floppy drive?
           + Q9.8: How do I run MATLAB in batch mode?
      _________________________________________________________________
  
MATLAB FAQ
**********
  
Introduction
************
  
    This is a list of frequently asked questions (with answers)
pertaining to MATLAB.  The current version of MATLAB at the time this
document was written is 6.5 (R13).
  
    This FAQ is freely redistributable.  I take no liability for the
correctness and safety of any procedures or advice given here.  This FAQ
is distributed in the hope that it will be useful, but WITHOUT ANY
WARRANTY; without even the implied warranty of MERCHANTABILITY or
FITNESS FOR A PARTICULAR PURPOSE.  Further, any opinions contained
herein are my own and not necessarily shared by my employer.
  
    I do require that if you redistribute this document, you do so in
unmodified form.  This simply helps ensure it will remain up-to-date.
If you are interested in a different formatting, check with me first, as
it may be easy to produce from the texinfo source.
  
    This FAQ is independent of and does not represent The MathWorks in
any way. The MathWorks does not officially sponsor, approve, or endorse
this FAQ or its content.
  
    The official and most recent hypertext version of this document can
be found at <http://www.mit.edu/~pwb/cssm/>;.
  
1 Introduction, Policy, Credits
*******************************
  
Q1.1: What is MATLAB?
=====================
  
    MATLAB is a commercial software package written by The Mathworks
(http://www.mathworks.com.  Quoting from their web page:
  
      Numeric computation, technical graphics and visualization, and an
      intuitive programming language for applications in engineering and
      science
  
    MATLAB is a complete environment for high-level programming, as well
as interactive data analysis.  MATLAB excels at numerical computations,
especially when dealing with vectors or matrices of data.  Symbolic math
is available through an add-on toolbox that uses a Maple kernel.
  
    There are too many toolbox add-ons to describe here.  Poke around the
web site if you have some interest.
  
    Questions about the name "MATLAB" often arise.  MATLAB stands for
"MATrix LABoratory".  See
<http://www.mathworks.com/access/ ... rn_matlab/ch1intro.
shtml>
for an overiew of MATLAB.
  
Q1.2: What is this newsgroup about?
===================================
  
    The newsgroup `comp.soft-sys.matlab' (sometimes abbreviated cssm) is
a forum for discussing issues related to the use of MATLAB.  It
occasionally includes questions related to similar software packages
like Octave.  Any topic related to MATLAB is appropriate.  Additionally,
there will be occasional discussions regarding related math topics in a
more abstract form.
  
    The original charter for the group, created in early 1993, can be
found at
<http://www.landfield.com/usenet/ ... omp/comp.soft-sys.m
atlab>.
  
    Before posting, please skim through this document to see if your
question has already been answered.  If it is has not, there may be
information here that may help you better understand the issue and
phrase your question.
  
Q1.3: Where can I find an archive of comp.soft-sys.matlab?
==========================================================
  
    Try the standard newsgroup archive at <http://groups.google.com>.
  
    Many people are also now using the web-based newsreader provided by
the MathWorks at their website, which contains archives back to
mid-2000.  It provides full posting capability.  See
<http://www.mathworks.com/matlabcentral/>;.
  
    There is also a complete, searchable archive of cssm at
<http://mathforum.org/epigone/comp.soft-sys.matlab/>;. It probably
contains the earliest articles of these services.
  
Q1.4: What other software packages exist to do similar work?
============================================================
  
    GNU Octave is a freely availabe software package with a language
"mostly compatible with MATLAB": <http://www.octave.org>.
  
    Scilab "is a scientific software package for numerical computations
in a user-friendly environment".  It is fully open source and has a
parallel version. <http://www-rocq.inria.fr/scilab/>;.
  
    IDL (Interactive Data Language) is a commericial software package
with applications similar to MATLAB.  It is very well suited to image
processing and 3D visualization.  IDL is produced by Research Systems
Inc., <http://www.rsinc.com>.
  
    O-Matrix is a commercial MATLAB-like program.  In fact it has a
MATLAB compatibility mode, which the authors claim can execute native
MATLAB code 5-10 times faster than MATLAB.  Readers who have used this
package are encouraged to send me a more detailed explanation.
<http://www.omatrix.com>
  
    LyME runs a reasonable subset of MATLAB code on the Palm platform.
Available for free at <http://www.calerga.com>. Thanks to Martin Cohen
for this info.
  
    Fredrik Hekland suggests: Back in the days when I used OS/2, Euler
was a good replacement for Matlab (at least for the simple operations I
needed at that time). I see that Euler is still living, now as GNU
GPL'ed OSS. <http://mathsrv.ku-eichstaett.de/MGF/homes/grothman/euler/>;
  
    Stefan Mueller is involved in developing a MATLAB-like program called
JMathLib, written in Java.  See <http://mathlib.sourceforge.net/>;.
  
Q1.5: What is the FAQ editorial policy?
=======================================
  
    This FAQ is actively maintained and should be up to date.  All
submissions are welcome; please email submissions to the maintainer:
<boettcher@ll.mit.edu>.  Submissions could be new questions (preferably
with answers), fixes such as dead links or typos, extra information to
elaborate on an answer, or a replacement answer.
  
    Questions and answers will be edited, and will be attributed.
Anything appearing without attribution was written by the FAQ
maintainer or by someone listed in the contributor section, unless it
is explicitly indicated as unknown or anonymous source.  Where answers
are quoted from Usenet news articles, they will always be attributed.
  
Q1.6: Who maintains this FAQ?
=============================
  
    This FAQ was created by Peter Boettcher <boettcher@ll.mit.edu> in
early 2001.  He is also the current maintainer.
  
Q1.7: Contributors
==================
  
    Thanks to Denis Gilbert for a substantial set of questions and
answers, as well as feedback on existing answers.  Steve Lord
<slord@mathworks.com> has contributed a number of additions and
refinements.  Many others have contributed suggestions and answers,
which are individually attributed.  Thanks!
  
  2 Basics
********
  
Q2.1: Why does MATLAB only calculate to 4 significant digits?
=============================================================
  
    It doesn't.  It uses full double-precision floating point numbers to
calculate everything.  By default it only prints a few decimal places
to the screen.  You can change this using the command `format long'.
Type `help format' for more information.
  
Q2.2: Why does the transpose operator take the complex conjugate?
=================================================================
  
    When performing linear algebra operations on complex matrices, it is
almost always the complex conjugate transpose (also called the Hermitian
transpose) that is needed (see Gilbert Strang's linear algebra book for
discussion- page 293 in edition 3).  The bare apostrophe is an operator
that takes the complex conjugate transpose.  The non-conjugating
transpose operator is a period followed by an apostrophe.  Type `help
punct' for more info.
  
      >> A'     % complex conjugate transpose
      >> A.'    % transpose
  
Q2.3: How can I set all NaN's to 0?
===================================
  
    NaN's are tricky.  `NaN==NaN' is false, so you can't say `x(x==NaN)
= 0;' Instead, use the function `isnan', as in: `x(isnan(x)) = 0;'
  
Q2.4: How can I make MATLAB use the full window width for displaying  
matrices?
=============================================================================
=
  
    Bob Gilmore <bgilmore@mathworks.com> writes:
  
      In R12 (MATLAB 6.0), this can be controlled via a preference.
      Select the `File | Preferences...' menu item, and select `Command
      Window' in the Preferences dialog that appears.  In the `Display'
      section, there's a checkbox labeled `Limit matrix display width to
      eighty columns.'  Unchecking that box allows matrix displays to
      make full use of the Command Window's width.  [Unchecked is the
      default.]
  
    Starting with MATLAB R12.1, users can access the current command
window size using the root property `CommandWindowSize'.  That is,
`sz=get(0, 'CommandWindowSize')'.  In R12.0, there is no way to do this
unless you call undocumented C functions from a MEX file.
  
Q2.5: How do I comment out a large block of code?
=================================================
  
    The built-in editor in MATLAB 6.0 now has a block-comment feature.
Or you can use matlab-mode
(http://www.mathworks.com/support/ftp/emacs_add_ons.shtml for Emacs,
which supports this as well.
  
    If you are using an older version, use this:
  
      if 0
         commented out code
      end
  
    This is not the best solution, since parse errors inside that block
will cause an error.  But it's better than sticking comment characters
in from of every line.
  
Q2.6: How do I save default settings across sessions?
=====================================================
  
    The key is to create a STARTUP.M file.  Look at the online help for
more detailed instructions specific to your operating system.
  
Q2.7: How can I find local maxima in a vector array?
====================================================
  
    You can use the following one-line function to determine the indices
of the local maxima.
  
      function index = localmax(x)
      index = find( diff( sign( diff([0; x(; 0]) ) ) < 0 );
  
Q2.8: Why is 6*i not complex in my program?
===========================================
  
    You might have overwritten the variable i with a number earlier in
your program (or session, for that matter).  You can reset i using
i=sqrt(-1), or "clear i", or you can use j, or you can use the function
"complex".
  
    Steve Lord <slord@mathworks.com> tells me that an even better
solution is to use 6i, which is unambiguously 6*sqrt(-1).  In MATLAB
versions >= 6.5, this can change a function which does not accelerate
to one which does.  For more details read
<http://www.mathworks.com/access/ ... lab_prog/ch7_per7.s
html>.
  
Q2.9: How can I modify the MATLAB path?
=======================================
  
    Easiest solution: use the PATHTOOL GUI.  Or if you want command line
access:
  
    Suggested by Joshua Stiff <Jcstiff@aol.com>: You can use `addpath'
to add directories from the command line, and `path2rc' to write the
current path back to `pathdef.m'.  If you do not have write permissions
for `pathdef.m', `path2rc' can write to a different file, which you can
execute from your `startup.m'.
  
  3 Programming
*************
  
Q3.1: Can I read a text file with a certain format?
===================================================
  
    Yes.  If the file has nothing but numbers separated by whitespace,
and has a constant number of columns through the entire file, you can
just type `load myfile.txt'.
  
    The function `DLMREAD' is more flexible and allows you to read files
with fields delimited by any character.
  
    The function `TEXTREAD' is more flexible still and allows you to skip
lines at the beginning, ignore certain comment lines, read text as well
as numbers, and more.  Type `help textread' for more info.
  
    If none of these suit your needs, you can always use the low-level
file I/O functions `FOPEN', `FREAD', `FSCANF', `FGETL', `FSEEK' and
`FCLOSE' to read the data however you would like.
  
Q3.2: Why are global variables bad?
===================================
  
    us <us@neurol.unizh.ch> writes in
<http://groups.google.com/groups?selm=90mer5$1u8$1@nnrp1.deja.com>
(lightly edited):
  
      I'll always fondly remember the outburst of the notorious and
      seasoned user Lars Gregersen to a recent poster (that most likely
      started this question) and will give it on to my students:
  
      The short answer is: Don't use global variables.
  
      The longer answer is: Don't use global variables unless you
      absolutely have to.
  
      Now,
  
      1) Using globals is perfectly ok (that's why they here on first
      place), just like `for''s and `while''s and other intelligible 2nd
      generation computer language constructs (that MATLAB is built on).
  
      2) Using globals is a blasphemy from a programmer's point of view
      because it shows that she/he didn't think ahead but was rather
      sloppy... and just kept adding on those never-ending "important
      set up" parameters that she/he needed to use by all the axes - or
      whatever - of a one single project ...
  
      3) Using globals is a problem in terms of book-keeping if(f) you
      end up having zillions of them hovering around in your workspace,
      and start having problems because one <my_par> from func1 is mixed
      up with <my_par> from func2 ... and <whos global> shows you 20
      pages worth of variables ...
  
      4) Using globals won't work in certain ML contexts, such as gui
      callbacks
  
      [Respectfully edited]
  
      Hence, eventually collecting your globals into a struct (i.e.,
      cleaning up your toolbox) and passing it on as a parameter of your
      functions is the most economic way to program in MATLAB ... and it
      shows your intelligence, to boot.
  
Q3.3: What about this logical array business?
=============================================
  
    From the _Getting Started_ book:
  
      The logical vectors created from logical and relational operations
      can be used to reference subarrays. Suppose X is an ordinary
      matrix and L is a matrix of the same size that is the result of
      some logical operation.  Then X(L) specifies the elements of X
      where the elements of L are nonzero.
  
    To remove the logical flag from an array, don't add 0 as suggested by
`help logical' in MATLAB 5.x.  This actually adds zero to each element
of the matrix.  Instead, use a `+' in front, as in `+x'.  This is a
solution given in `help logical' in MATLAB 6.
  
Q3.4: Huge memory waste using array of structs?
===============================================
  
    The following example was posted to the newsgroup:
  
      I've discovered to my horror that structs take up an obscene
      amount of overhead (I'm running version 5.3.1.29215a (R11.1) on a
      Dec ALPHA). I have a set of 10,242 observations, each consisting
      of 3+13=16 fields, which have 3*27 + 1*13 = 94 values.  So the
      total size in bytes should be 10,242 * 94 * 8 bytes/double =
      7,701,984.
  
      I have this stored in a 1 x 10242 data structure, and when I issue
      the `whos' command, it tells me that the data now takes up
      27,367,136 bytes!
  
    Cris Luengo <cris@ph.tn.tudelft.nl> answers:
  
      My guess would be that a structure contains MATLAB arrays. Each
      array has some overhead, like data type, array sizes, etc. In your
      second implementation (index using data.latitude(observation)),
      there are 10,242 times less arrays allocated. Note that in your
      data, for each observation, you have 13 arrays with one value. I
      don't know how large the matrix header exactly is, but it is a
      waste putting only a single value in it!
  
    I think Cris has hit it exactly.  Every MATLAB matrix has an
overhead of ~100 bytes, even matrices with a single element.  In this
example, there are 16 fields * 10242 elements = 163872 matrices.  Each
one of these matrices adds an additional 100 bytes, for 16.4Mbytes in
pure overhead.  This still comes up a little short of the amount
reported, but it is fairly close.
  
    It is much more efficient, both for storage and computation, to use a
struct of arrays rather than an array of structs.
  
Q3.5: Why is my MEX file crashing?
==================================
  
    Memory errors are one likely reason.  Greg Wolodkin
<greg@fess.mathworks.com> suggests the debug memory manager:
  
      The platform-independent way to use the debug memory manager is to
      set the environment variable MATLAB_MEM_MGR to contain the string
      "debug".
  
           On Windows:
            
             C:\> set MATLAB_MEM_MGR=debug
             C:\> matlab
            
           On Unix with csh or tcsh:
            
             % setenv MATLAB_MEM_MGR debug
             % matlab
            
           On Unix with sh or bash:
            
             $ MATLAB_MEM_MGR=debug matlab
  
      The debug memory manager cannot catch your code the instant it
      writes out of bounds (tools like Purify *can* do this but the
      performance hit they induce is quite painful).  What it will catch
      is that in general, when you write outside of one memory block you
      end up writing into another, corrupting it or (in the case of the
      debug memory manager) hopefully corrupting only a guard band.
      When you later free the memory, we can tell you that you walked
      off the end of the block and corrupted the guard band.
  
Q3.6: How can I create variables A1, A2,...,A10 in a loop?
==========================================================
  
    Don't do this.  You will find that MATLAB arrays (either numeric or
cell) will let you do the same thing in a much faster, much more
readable way.  For example, if A1 through A10 contain scalars, use:
  
      A = zeros(1,10);        % Not necessary, just much faster
      for i=1:10
        A(i) = % some equation
      end
  
    Now refer to `A(i)' whenever you mean Ai.  In case each Ai contains
a vector or matrix, each with a different size, you want to use cell
arrays, which are intended exactly for this:
  
      for i=1:10
        A{i} = 1:i;
      end
  
    Note that each `A{i}' contains a different size matrix.  And be
careful to use the curly braces for the subscript!
  
    Now, if you still *really* want to create variables with dynamically
generated names, you need to use `eval'.  With eval, you use MATLAB
commands to generate the string that will perform the operation you
intend.  For example, `eval('A=10')' has the same effect as `A=10', and
`eval(['A' 'B' '=10'])' has the same effect as `AB=10', only the `eval'
method executes much more slowly.  So in a loop, you could use:
  
      for i=1:10
        eval(sprintf('A%d = [1:i]', i));
      end
  
    Notice how much more obfuscated this is.  Repeat: don't do this
unless you have a very good reason (such as someone gives you a MAT
file with 2000 variables named A1428, for example).
  
Q3.7: Do boolean operators short-circuit?
=========================================
  
    In many programming languages, boolean operators like AND and OR will
stop evaluating as soon as the result is known.  For instance,
  
      1 | error('Short-circuit')
  
    would never get to the error part, since the 1 is always true.
  
    MATLAB versions >= 6.5 include the new short-circuiting logical
operators || and &&.  Use these for all condition tests in loops and
similar, and use the old | and & for element-by-element logical
operations.  You can find details at
<http://www.mathworks.com/access/ ... es/matlab/matlab134
.shtml#53717>
  
    In older versions of MATLAB, the boolean operators | and & are only
short-circuit evaluated inside the conditions of IF and WHILE
statements.  In all other contexts, all parts of the conditional are
evaluated.  For an interesting discussion on this topic, see
  
<http://groups.google.com/groups? ... f@frenchslinux.dhcp>;
  
Q3.8: How do I fix "Out of Memory" problems?
============================================
  
    A frequent variant of this question is: "I have 512M of RAM, and 2G
of swap space.  Why can't I create this 200M matrix?"
  
    Simple answers first: Remember that double precision floats take up 8
bytes.  So a million element vector takes up 8Mbytes.  Be sure you're
estimating properly.
  
    Many operations need to create duplicate matrices.  For example,
B=inv(A.')  must create a tempory variable the same size as A to hold
the transpose, and B is again, the same size as A.
  
    If you're sure your matrices are reasonably sized, then read all of
TMW Tech Note 1106, a great reference:
<http://www.mathworks.com/support/tech-notes/1100/1106.shtml>;
  
Q3.9: How do I dynamically generate a filename for SAVE?
========================================================
  
    You're probably trying
  
      fname = 'foobag';
      save fname variable;
  
    To do this correctly, you need to use the "functional" form of save:
  
      fname = 'foobar';
      save(fname, 'variable');
  
    In fact, it is true in general that the following two lines are
equivalent:
  
      command str1 str2 str3
      command('str1', 'str2', 'str3')
  
    This allows one replace any or all of the parameters with dynamically
generated strings.  This is also useful in commands like PRINT, LOAD,
CLEAR, etc.
  
Q3.10: What's the difference between M-files, Pcode, and MEX files?
===================================================================
  
    Suggested by Joshua Stiff:
  
    * M-files are plain ASCII text that is interpreted at run time.
      Actually it is parsed once and "just-in-time" compiled, but this is
      transparent to the user.  Use M-files for most of your MATLAB
      development, and for platform independence and maintainability.
  
    * Pcode is a preparsed and encoded version of the M-file.  Since it
      is preparsed, it saves on the load time of the function.  This is
      most likely not an issue except for very large M-files, since most
      are parsed only once anyway.  Pcode also lets you hide the source
      code from others.  Careful, there is no way to convert Pcode back
      to the M-file source.  Pcode is platform independent.
  
    * MEX files are native C or C++ files that are dynamically linked
      directly into the MATLAB application at runtime.  They must be
      compiled for each hardware architecture on which they are to be
      run.  MEX files have the potential to crash the MATLAB
      application, but rather large speed gains are possible, depending
      on the algorithm.
  
Q3.11: Can MATLAB pass by reference?
====================================
  
    This is really two questions in one.  One is "Can I modify a
function's input argument?"  This would save memory and simplify
programming in some cases.  The answer here is "NO".  If you modify the
input argument of a function, all you do is modify a copy of the
argument local to the function.  The only way to modify variables from
a function is to return the result when finished, as in
  
      bigstruct = addelement(bigstruct, 5);
  
    The other question is: "Pass by value wastes memory and time, since
copies of variables are made.  How can I fix this?"  Here, the answer
is "Your assumption is flawed, you don't need to."  MATLAB uses a
scheme called "copy-on-write" to optimize this sort of thing.
Basically, data is shared between variables whenever possible, and a
true copy is made only when one of the variables is modified.  So
although MATLAB's calling convention appears to be pass-by-value, if
you don't modify the input variables, the data is never copied.
  
Q3.12: How can I process a sequence of files?
=============================================
  
    If you can generate the filename using an incrementing counter, use
code like this:
  
      for k=1:20
        fname=sprintf('/path-name/m%d.dat',k);
      
        data=load(fname);
        % or
        data=imread(fname);
        % or
        fid=fopen(fname, 'rb');
        fread(fid, ...);
      end
  
    If instead you want to process all the files in a directory, you
might instead wish to use `dir':
  
      d=dir('*.jpg');
      for k=1:length(d)
        fname=d(k).name;
        % ...
      end
  
  4 Graphics
**********
  
Q4.1: How do I adjust the fontsize of a ticklabel?
==================================================
  
    The ticklabel gets its properties from the axis to which it is
attached.  So `set(gca, 'fontsize', 14)' should do the trick.  Type
`get(gca)' to see what else you can set.
  
Q4.2: Can I create a pi/sigma/superscript in my ticklabels?
===========================================================
  
    Not directly... MATLAB does not interpret TeX strings in ticklabels.
You can play games with placing text by hand.  The following MathWorks
solutions may give you some guidance:
  
    <http://www.mathworks.com/support/solutions/data/27450.shtml>;
  
    and
  
    <http://www.mathworks.com/support/solutions/data/5375.shtml>;
  
    There are also some free third-party software packages you can use to
accomplish this.
  
    Doug Schwarz has written a Styled Text Toolbox that does this.  It is
freely available at:
<http://home.earthlink.net/~dschwarz2/stextfun/index.html>;
  
Q4.3: Can I open multiple files using uigetfile?
================================================
  
    No, but there is file called `uigetfiles.dll' on The MathWorks web
site that will allow you to do this, at least on a Windows platform.  It
can be found on this page:
<http://www.mathworks.com/matlabc ... .jsp?type=category&
id=&fileId=331>
  
Q4.4: I want to use a scrollbar to scroll my edit boxes/buttons
===============================================================
  
    Sorry, there's no easy solution.  MATLAB does not support
hierarchical figures, so you can't have a container control holding
your controls.  If you really need this you'll have to create your own,
using the callbacks from the scrollbar to modify the position
properties of your controls.
  
    Ghassan Hamarneh <ghassan@s2.chalmers.se> writes:
  
      What I would do is to add 2 pushbuttons to the figure: one at the
      top right and another at the bottom right and use these buttons to
      control the vertical scroll of the content of the figure. (surely
      you can add another 2 horizontal pushbuttons, lower left and lower
      right).
  
      Whenever any of these buttons is pressed, you loop over all the
      controls except the two pushbuttons, and increment/decrement the
      vertical/horizontal postition value of each control. Something
      like this:
  
             % fig_handle is the handle of the figure containing the UI  
controls
             % pushb1_handle, pushb2_handle  are the handles of the  
pushbuttons
             % described above
            
             % find handles of all the controls
             all_handles=findobj(fig_handle);
            
             % exclude the handles of the 2 pushbuttons and the figure itself
             move_handles=setdiff(all_handles, ...
                                  [fig_handle, pushb1_handle,  
pushb2_handle]);
            
             % loop over the remaining handles and change their positions
             for k=1:length(move_handles)
               set(move_handles(k),'position', ...
                                 (get(move_handles(k),'position'))-[0 0  0  
10]);
             end
  
Q4.5: How can I rotate ticklabels?
==================================
  
    You can't do that directly.  Use the following Mathworks solutions to
place the labels manually as text objects (this is a hack):
  
    <http://www.mathworks.com/support/solutions/data/5375.shtml>;
  
Q4.6: How can I display data in a grid like Excel?
==================================================
  
    If you simply want to edit the matrix as if it were an Excel
spreadsheet, you can use the builtin array editor in MATLAB 6.  Type
`openvar(my_var)' at the command prompt, or double click on the
variable in the workspace browser.  This editor is limited to 2-D
matrices of less than 10000 elements.
  
    If you are working on a PC, an ActiveX object is the way to go.  Try
searching the archive for "activex" and "grid".  Michael Robbins has
posted a number of articles on this topic, so you might try adding his
name to the search.  Try a search like this:
  
<http://groups.google.com/advance ... comp.soft-sys.matla
b&as_uauthors=robbins&as_q=activex%20grid>
  
Q4.7: How can I write a title or text with multiple lines?
==========================================================
  
    This can be done using a cell array of strings:
  
            title({'First line','Second Line'})
            text(0.5,0.5,{'First line','Second Line'})
  
Q4.8: How can I use a common color scale among several images?
==============================================================
  
    One way to do this is to insert a CAXIS command in your plotting
function/script.  For this to work well, you first need to have a first
look at all of your data to determine what are the minimum and maximum
values over the entire set of images.   For example, if the overall
minimum value amongst all images is 40 and the overall maximum is 256,
you may issue the command "caxis([40 260])" for each image.
  
    Tech note 1215, at
<http://www.mathworks.com/support/tech-notes/1200/1215.shtml>;
addresses a related question, namely "How can I use multiple colormaps
in a single figure".  As a bonus, it includes a thorough discussion of
colormaps in general.
  
Q4.9: How can I set default handle graphics properties?
=======================================================
  
    There are probably several hundred of these default handle graphics
options.  Rather than trying to remember any particular one, the best
thing is to learn the general principle behind all of these default
handle graphics properties.  The basic call to insert into your
startup.m file is :
  
            set(0,'DefaultObjectnamePropertyName',Value)
  
    For line objects, here are a few examples:
  
      set(0,'DefaultLineMarkerSize',12);
      set(0,'DefaultLineMarker','d');
      set(0,'DefaultLineLineWidth', 2);
  
    Similarly, you can use these statements for axes objects:
  
      set(0,'DefaultAxesLineWidth', 2);
      set(0,'DefaultAxesXGrid','on');
      set(0,'DefaultAxesTickDir','out');
      set(0,'DefaultAxesTickLength',[0.015 0.015]);
      set(0,'DefaultAxesFontName','Arial')
  
    For more details, do a full text search for 'Defining Default
Values' in the R12 online help, and click on the very first hit.  Also
see the following entries in the R12 online help:
  
    * Graphics Object Hierarchy
  
    * Types of Graphics Objects
  
Q4.10: How can I modify the default background color used for histograms?
=========================================================================
  
    A histogram is made up of patch objects.  The trick is to modify the
FaceColor property of these patches.  A short example follows:
  
      x=rand(400,1);
      hist(x);                           % Default facecolor is blue
      h=get(gca,'Children');
      set(h,'FaceColor', 'm');           % magenta facecolor
      
      % Draw only the edges of the bars making up the histogram
      set(h,'FaceColor', 'none');
  
Q4.11: How can I draw more than two lines with plotyy?
======================================================
  
    You can use the axes' handles to plot additional lines, as follows:
  
      x1 = (1:10)';  x2 = x1;  y1 = x1; y2 = x1.^2;
      %Plot one line against each of the left and right y-axis
      [ax, h1, h2] = plotyy(x1,y1,x2,y2);
      
      %Plot additional line against the left y-axis
      x3= x1;  y3 = y1 + 3;
      h3 = line(x3,y3,'Parent', ax(1), 'Color',get(h1,'Color'));
      
      %Plot additional line against the right y-axis
      x4= x1;  y4 = y2+3;
      h4 = line(x4,y4,'Parent', ax(2), 'Color',get(h2,'Color'));
      set(ax,'YLimMode','auto','YTickMode','auto')
  
Q4.12: Is there a command available for plotting an arrow on a graph?
=====================================================================
  
    A user-contributed m-file (arrow.m) is available at
<http://www.mathworks.com/matlabc ... .jsp?type=category&
id=&fileId=278>
  
    Also look at arrow3.m at
<http://www.mathworks.com/matlabc ... .jsp?type=category&
id=36&fileId=1430>
  
    Ned Gulley of The MathWorks has written a review of both of these
utilities at
<http://www.mathworks.com/matlabcentral/spotlight/arrows.shtml>;
  
Q4.13: Why does movie(M,1) display the movie M twice?
=====================================================
  
    Quoting from the Matlab 6.1 online helpdesk: The movie function
displays each frame as it loads the data into memory, and then plays
the movie. This eliminates long delays with a blank screen when you
load a memory-intensive movie. The movie's load cycle is not considered
one of the movie repetitions.
  
Q4.14: How can I set the focus in my GUI?
=========================================
  
    You can't.  One hopes that The MathWorks will include this
often-requested feature in a future, but there is no guarantee.
  
    Related to this, changing the stacking order of your GUI elements
_might_ allow you to set the tab order, but this seems to not always
work.  GUIDE in MATLAB version >= 6.5 includes a Tab Order Editor,
which does a better job at this.
  
Q4.15: How can I add text labels to data points?
================================================
  
    The `text' command can be used in a vectorized form to automatically
add text labels wherever needed.  Say you have a matrix D, where the
first column contains X coordinates and the second column contains Y
coordinates.  Then us <us@neurol.unizh.ch> illustrates:
  
      plot(D(:,1),D(:,2),'+-');
      n=num2str(D,'%5.3f/');
      n=n(:,1:end-1);          % Just to remove the trailing slash
      text(D(:,1),D(:,2),n);
  
  5 Math/Algorithms
*****************
  
Q5.1: Why is 0.3-0.2-0.1 not equal to zero (or similar)?
========================================================
  
    (The above example returns -2.7756e-17.)  Jason Bowman
<jbowman90@hotmail.com> writes:
  
      As is mentioned all the time in this newsgroup, some floating
      point numbers can not be represented exactly in binary form.  So
      that's why you see the very small but not zero result.  See `EPS'.
  
      [...]
  
      The difference is that `0:0.1:0.4' increments by a number very
      close to but not exactly 0.1 for the reasons mentioned below.  So
      after a few steps it will be off whereas `[0 0.1 0.2 0.3 0.4]' is
      forcing the the numbers to their proper value, as accurately as
      they can be represented anyway.
  
    us <us@neurol.unizh.ch> elaborates:
  
      Watch Jason's response unfolding:
           a=[0 0.1 0.2 0.3 0.4];
           b=[0:.1:.4];
            
           sprintf('%20.18f\n',a)
           ans =
           0.000000000000000000
           0.100000000000000010
           0.200000000000000010
           0.299999999999999990
           0.400000000000000020
            
           sprintf('%20.18f\n',b)
           ans =
           0.000000000000000000
           0.100000000000000010
           0.200000000000000010
           0.300000000000000040
           0.400000000000000020
  
    For more detail then you ever will want on floating point arithmetic,
skim the following paper: "What Every Computer Scientist Should Know
About Floating Point Arithmetic":
<http://www.validlab.com/goldberg/paper.ps>;
  
    Also take a look at the wonderful Cleve's Corner article from 1996.
It is much more accessible than the previous link.
<http://www.mathworks.com/company/newsletter/pdf/Fall96Cleve.pdf>;
  
Q5.2: How does the backslash operator work?  What does it do?
=============================================================
  
    For full matrices, pseudocode describing the algorithm can be found
here:
  
    <http://www.mathworks.com/support/solutions/data/22317.shtml>;
  
    For sparse matrices, the URL is:
<http://www.mathworks.com/support/solutions/data/21798.shtml>;
  
    Also for sparse matrices, you can turn on monitoring routines that
show you some of the steps.  Use `spparms('spumoni', 1)', or
`spparms('spumoni', 2)'
  
Q5.3: How does one compute a factorial?
=======================================
  
    After MATLAB 5.3, there is a `factorial' function, but it is not
vectorized.  Why this is so we will never know.  Instead, you can use
the `gamma' function, which is vectorized.  Careful, `factorial(n) =
gamma(n+1)'.  If you are trying to compute a ratio of factorials, see
the next question.
  
Q5.4: How can one accurately compute ratios of factorials?
==========================================================
  
    If you are trying to compute "n choose k", just use the function
`nchoosek'.  Otherwise, Paul Skoczylas <pauls@cfertech.com> suggests in
<http://groups.google.com/groups?selm=bzAM5.2437$pQ4.21981@jekyl.ab.tac.net>
  
      If I wanted to evaluate n!/(n-j)! for large values of n and/or j
      (but still assuming n>j), I would use the `gammaln' function.
  
           gamma(m+1)=m!
           gammaln(m+1)=log(m!)
  
      Rewrite and manipulate the equation:
  
           A=n!/(n-j)!
           log(A)=log(n!/(n-j)!)
           log(A)=log(n!)-log((n-j)!)
           A=exp(log(n!)-log((n-j)!))
      so,
  
           A=exp(gammaln(n+1)-gammaln(n-j+1))
  
Q5.5: How can I fit a circle to a set of XY data?
=================================================
  
    An elegant chunk of code to perform least-squares circle fitting was
written by Bucher Izhak and has been floating around the newgroup for
some time.  The first reference to it that I can find is in
<http://groups.google.com/groups? ... 32886D%40skm.com.au>;:
  
      function [xc,yc,R,a] = circfit(x,y)
      %CIRCFIT  Fits a circle in x,y plane
      %
      % [XC, YC, R, A] = CIRCFIT(X,Y)
      % Result is center point (yc,xc) and radius R.  A is an optional
      % output describing the circle's equation:
      %
      %   x^2+y^2+a(1)*x+a(2)*y+a(3)=0
      
      % by Bucher izhak 25/oct/1991
      
      n=length(x);  xx=x.*x; yy=y.*y; xy=x.*y;
      A=[sum(x) sum(y) n;sum(xy) sum(yy) sum(y);sum(xx) sum(xy) sum(x)];
      B=[-sum(xx+yy) ; -sum(xx.*y+yy.*y) ; -sum(xx.*x+xy.*y)];
      a=A\B;
      xc = -.5*a(1);
      yc = -.5*a(2);
      R  =  sqrt((a(1)^2+a(2)^2)/4-a(3));
  
    Tom Davis provided a more sophisticated approach that works for more
cases in
<http://groups.google.com/groups? ... 50DA497@eng.usf.edu>;and
<http://groups.google.com/groups? ... 7F192BC@eng.usf.edu>;.
Code included.
  
Q5.6: Why does MATLAB return an complex number for (-8)^(1/3)
=============================================================
  
    In the same way there are two solutions (plus and minus) for the
square root of a positive number, there are multiple solutions for
roots of negative (and complex) numbers.  If you express the number in
magnitude*exp(i*theta) form, the cube root (for instance) takes the
form (magnitude^(1/3))*exp(i*theta/3*k), for k=1:3.
  
    -8 is 8*exp(i*pi), so the cube roots are 2*exp(i*pi/3),
2*exp(2*i*pi/3), and 2*exp(3*i*pi/3).  The last one simplifies to -2.
  
    MATLAB always returns the _first_ solution counter-clockwise from the
positive real axis.  Armed with this knowledge, you can compute all or
some particular root.  For instance, if you want the negative real cube
root, simply take the cube root of the absolute value of the number,
and negate it.

   For a different wording and more information, see
<http://www.mathworks.com/support/solutions/data/2998.shtml>;.

   Finally, Joe Sababa <joesababa@yahoo.com> suggests in
<http://groups.google.com/groups? ... WebX.raydaftYaTP>; a
method to find all the roots at once:

     Find the roots of a polynomial:
      P=[1 0 0 27];
      roots(P)

Q5.7: Can I compute the DFT of an irregularly-sampled signal?
=============================================================

   Ken Davis <kendavis@verizon.net> writes:

     The easiest way to find the spectrum of irregularly sampled data
     is to resample it to uniform samples. You can do this with MATLAB's
     `interp1' function. The accuracy of your spectrum will depend on
     the accuracy of the interpolation. You will want to experiment with
     several of the interpolation methods that are available in
     `interp1'. I believe that for interpolation with a limited window
     (i.e. interpolating a sample value from N nearest neighbors), the
     Lagrange interpolation is optimal, but Lagrange is not one of the
     choices in `interp1'.

     If interpolation doesn't work there are other schemes available.
     The Lomb-Scargle periodogram is often mentioned in relation to
     this question and may be more appropriate if your data has very
     uneven spacing (e.g. very large or very small spacings). I know
     that this algorithm is listed in Numerical Recipes, but I don't
     have a good on-line reference (with MATLAB code) to point you to.

     In general, the problem is that the spacing between points
     determines the "importance" of a particular point. For example, if
     several points are very close together, then small amounts of
     noise on those measurements will tend to have a greater effect on
     inferring the slope of the function (and with it, the high
     frequency energy) than the same amounts of noise on measurements
     that are further apart.

Q5.8: How can I efficiently do row-wise vector-matrix multiplication?
=====================================================================

   Elaborating on the question, the for-loop implementation looks like:

     for row = 1:N
       output(row,  = vector(row) * matrix(row, ;
     end

   In recent MATLAB versions (>= 6.5), this is actually not such a bad
idea, as the accelerator will do a good job on a loop like this.

   The "old-style" vectorized version of this uses `repmat'.

     output = matrix .* repmat(vector, 1, M);

   This is a fine solution, except that having to replicate `vector'
uses memory unnecessarily, and is less cache-efficient.  The
accelerated single loop may run faster in some cases, due to the better
cache-usage.

   Finally, there is a solution using `sparse'.  Ken Davis
<kendavis@verizon.net> writes:

     If  A is MxN and B is  Nx1 or 1xN, then A*sparse(1:N, 1:N, B)
     multiplies the columns of A by the elements of B. Similarly,
     sparse(1:M, 1:M, B)*A multiplies the rows of A by the
     corresponding elements of B. For division, use 1./B in place of B
     as the argument to sparse.


   This solution requires the conversion of the full `vector' to a
sparse format, but the actual computation will be very fast.  The
fastest solution depends on the sizes involved, so try them all!

6 Installation, Crashes, Platform-Specific Issues
*************************************************

Q6.1: MATLAB 5.3 won't run on my linux system.
==============================================

   First try `matlab -n' to see which shared libraries MATLAB can or
can't find.  If you get a `"file not found"' error instead of a list of
shared libraries, you're missing `/lib/ld.so.1' or something similar.
Otherwise you will see the name of each library MATLAB needs and the
location of the one it found.  The big one is libc5.

   If you are missing `/lib/ld.so.1', you probably don't have the
correct version of the linux dynamic linker.  Your distribution likely
offers libc5 compatibility packages.

   Greg Wolodkin <greg@fess.mathworks.com> suggests:

     I would download 1.9.11 (the tar.gz file from the above site),
     untar it, and run the `instldso.sh' script therein.  Run it a
     second time using `--force' after you read the information
     displayed.

   After fixing the ld.so problem, I also had to install X libraries
which were compatible with libc5.  `www.xfree86.org' has binaries for
that kind of system... drop the lib files in a different directory
(`/usr/X11R6/lib-compat' or something), add this to `/etc/ld.so.conf',
and re-run `ldconfig'.

   Lars Jansson <lj@s2.chalmers.se> shares his easy Redhat solution,
using RPMs from an old version of Redhat:
     % rpm -i /Path_to_RedHat_6.2_CD/RedHat/RPMS/ld.so-1.9.5-13.i386.rpm
     % rpm -i /Path_to_RedHat_6.2_CD/RedHat/RPMS/libc-5.3.12-31.i386.rpm

   Don Chorman <chormand@bounce-msu.edu> has instructions for Suse 7.3,
from thread
<http://groups.google.com/groups?selm=aaime6$2cen$1@msunews.cl.msu.edu>:

   Before I installed Matlab, I installed the shlibs5(old version 5)
package. This package came on one of the installation CD's. I used the
Yast2 control center to install the package. Under Sofware - add/remove,
search the packages for shlibs to find the package) Package info:
     --------------------------------------------------------------------------
     Version: 2001.7.30-22, Thu Sep 20 00:13:43 2001
     Required by: shlibs5
     Description: Shared C libraries needed to run programs linked
     with libc and libm libraries (version 5).
     --------------------------------------------------------------------------

Q6.2: Why do I get a segfault after compiling correct MEX files under linux?
============================================================================

   Christian Joensson <c.christian.joensson@telia.com> answers:

     I suspect you have MATLAB 5.2.1 (R11.1) or prior and that you have
     a libc6 (glibc2) based linux system.  Either (1) upgrade to MATLAB
     6 (R12) or (2) downgrade to a libc5 based linux system or (3) use
     the cross-compiler mentioned at
     <http://www.mathworks.com/support/solutions/data/11129.shtml>;

   If you run a RedHat system, the solution above contains cookbook
instructions for installing a libc5 cross-compiler that will happily
co-exist with your regular compiler.  If you run a different
distribution, use the above solution as a guide, but you will need to
do some hacking to get things all set up.  If you can provide a
detailed fix for a specific system other than the one mentioned here,
please email the maintainer of this FAQ!

Q6.3: Why does MATLAB 6.0 crash on my Pentium 4?
================================================

   If the error message you receive is "Failure Loading Desktop Class",
see solution 27293 at The MathWorks:
<http://www.mathworks.com/support/solutions/data/27293.shtml>;

Q6.4: How can I make MATLAB 6.0 (R12) open without the GUI?
===========================================================

   Start MATLAB using the command `matlab -nodesktop'.  A related
switch is `-nojvm', which starts MATLAB without the Java Virtual
Machine, making it take much less memory.  However many of the editor
and browser features will not work.

Q6.5: Why does MATLAB 6.0 crash under Redhat 7.0?
=================================================

   Greg Wolodkin <greg@fess.mathworks.com> writes in
<http://groups.google.com/groups?selm=94msc5$61m$1@news.mathworks.com>:

     This is a well-known bug in the version of glibc (2.1.92) that
     Redhat chose to ship with Redhat 7.0.  Any program which uses
     LinuxThreads is pretty much guaranteed not to work.  The topic has
     come up in this newsgroup, though not recently.

     We mention the Redhat 7 problems in the release notes.  There is
     also a tech support SWAT available at
     <http://www.mathworks.com/support'; search for `redhat 7> and it
     will be the only hit.

     <http://www.mathworks.com/support/solutions/data/26841.shtml>;

     The fix is to upgrade to glibc 2.1.94 or better.  I'd recommend
     glibc 2.2 at this point.

     A short-term work-around is to start using `matlab -nojvm' which
     makes MATLAB a single-threaded program.

Q6.6: Why can't I type into the command window?
===============================================

   Greg Wolodkin <greg@fess.mathworks.com> writes in
<http://groups.google.com/groups?selm=9de9hp$avr$1@news.mathworks.com>:

     If you are running XFree86 4.0, try turning off your num lock key.
     Or upgrade to the very latest XFree86 4.0.x where I believe that
     bug is fixed.

     As a short-term workaround you can use "matlab -nojvm" to get back
     to the R11 xterm interface, which doesn't have this problem.

Q6.7: What happened to MATLAB for the Macintosh?
================================================

   It's back!  Development ceased at version 5.2.1, but has now resumed
with 6.5 for OSX.  As it is the first OSX version there are reportedly
a few quirks but it seems to be running quite well in general.

   Read about it: <http://www.mathworks.com/programs/mac/index.shtml>;

Q6.8: How do I release a MATLAB license?
========================================

   The MathWorks has prepared an answer to this question at:

   <http://www.mathworks.com/support/solutions/data/976.shtml>;

Q6.9: Is MATLAB optimized for the Pentium IV?
=============================================

   MATLAB 6.5 (Release 13) does indeed contain BLAS libraries optimized
for the Pentium IV, though more recent versions of ATLAS (the BLAS
library included with MATLAB) will further increase performance.

   If you are using Release 12 or want to try to improve performance,
you can compile your own.  Kevin Sheppard has been kind enough to
prepare a web page with instructions for compiling your own ATLAS
libraries on windows.  He also provides ready-to-download DLLs for most
common x86 processors.  Find it at
<http://weber.ucsd.edu/~ksheppar/matlab/MatlabAtlas.htm>;.

   If you are running Linux, search the newsgroup archives for
references to ATLAS.  It is straightforward to compile your own
customized version.

Q6.10: Why doesn't Ctrl-C interrupt execution of my m-file?
===========================================================

   This is likely to be a problem only under Windows, where MATLAB must
poll for Ctrl-C events.  If it is deep within matrix computation, it
will simply not respond.  If this occurs inside a loop construct, you
can force MATLAB to poll more often by inserting `drawnow' or
`pause(0)' into the loop.  This will also update your figures and make
GUIs more responsive.

7 Toolboxes, Software Archives, Books, and Other Resources
**********************************************************

Q7.1: Where can I find archives of MATLAB software?
===================================================

   The Mathworks has launched a new website for user-contributed MATLAB
files: <http://www.matlabcentral.com/>;  This URL replaces the old FTP
site.

   Another nice archive can be found at Mathtools.net, at
<http://www.mathtools.net/MATLAB/index.html>;

Q7.2: Where can I find a package to perform a specific task?
============================================================

   First, see the previous question for the first place to look.

   _Styled Text Toolbox_: written by Doug Schwarz, this toolbox allows
extremely flexible formatting of text strings, including symbols, math
formulae, etc.  Its TeX interpreter is much more complete than the
builtin MATLAB interpreter.
<http://www.servtech.com/~schwarz/stextfun/index.html>;

   _MATLAB and LaTeX_: Arno Linnemann has written an M-file to simplify
the inclusion of MATLAB graphics into LaTeX documents, along with a
nice document of tips and tricks.
<http://www.uni-kassel.de/~linne/matlab/WelcomeEng.html>;

   _Connectaxes_: Wolfgang Hammer has written a function that is used
in conjunction with zoom.  It allows zoom in one axes to control other
axes.  <http://www.lem.ee.ethz.ch/plecs/GOODIES/e_goodies_misc.html>;

   _Genetic Algorithm Optimization Toolbox_: GAOT implements simulated
evolution in the MATLAB environment.  Written by the Meta-Heuristic
Research and Applications Group at the North Carolina State University
Department of Industrial Engineering: <http://www.ie.ncsu.edu/mirage/>;

Q7.3: What textbook would you recommend for learning Matlab?
============================================================

   A popular textbook for a general audience is Mastering Matlab 6 by
Duane Hanselman and Bruce Littlefield
(<http://www.eece.maine.edu/mm/>;).  Other more specialized textbooks on
a variety of disciplines are listed on the Mathworks web site at

zzitmf

要是能翻译就好了