tag:www.nabble.com,2006:forum-1895Nabble - Octave2008-07-24T13:05:28ZGNU Octave is a high-level language, primarily intended for numerical computations. It provides a convenient command line interface for solving linear and nonlinear problems numerically, and for performing other numerical experiments using a language that is mostly compatible with Matlab. It may also be used as a batch-oriented language. Octave home is <a href="http://www.octave.org/" target="_top" rel="nofollow">here</a>.tag:www.nabble.com,2006:post-18639008Re: gnuplot zoom2008-07-24T13:05:28Z2008-07-24T13:05:28ZgOS<br><blockquote class="quote light-black dark-border-color"><div class="quote light-border-color">
<div class="quote-author" style="font-weight: bold;">Michael Goffioul-2 wrote:</div>
<div class="quote-message">I might be wrong, but I thought that mous zooming in gnuplot
<br>was only available in (not yet released) 4.3 version.
</div>
</div></blockquote>
I was looking through the mailing list and may have been mislead.
<br><br>It says <a href="http://www.nabble.com/zoom-to9221809.html#a9221809" target="_top">here</a> that as early as 2004 there were zoom capabilities.
<br><br>Then there was mention of using set mouse, which I think no longer works from another page.
<br><br>Then <a href="http://www.nabble.com/Mouse-zoom-and-keyborad-shortcuts-not-working-with-Octave-3.0-on-windows-XP-to15699262.html#a15709554" target="_top">this</a> suggested 4.3 which got me looking.
<br><br>Next I found the bit about being able to use cygwin's gnuplot <a href="http://www.nabble.com/cygwin-gnuplot-(CVS)-can-be-used-from-mingw-octave-3.0.0-(mouse-zoomin-active!!)-to15322278.html#a15322278" target="_top">...</a> but I didn't feel like installing cygwin again, especially since octave wouldn't even run properly or plot last time I tried it (About 2 weeks ago).
<br><br>Finally I saw this <a href="http://www.nabble.com/gnuplot-4.3.0-2008-03-24---mouse-zooming-works-to16814618.html#a16814618" target="_top">gnuplot 4.3 works.</a> So I downloaded the binaries and it didn't work for me, and I don't have a gcc compiler or the desire to install one atm, so I didn't try recompiling from source either.
<br><br>So I figured I'd through out a line and see if anyone had any information. It seems though that I have to wait till 4.3 comes out...
<br><br><blockquote class="quote light-black dark-border-color"><div class="quote light-border-color">
<div class="quote-author" style="font-weight: bold;">Michael Goffioul-2 wrote:</div>
<div class="quote-message shrinkable-quote">> Anyhow, I'm using the 3.0.1 (vs 2005) version of Octave for windows located
<br>> here:
<br>> <a href="http://sourceforge.net/project/showfiles.php?group_id=2888&package_id=40078" target="_top" rel="nofollow">http://sourceforge.net/project/showfiles.php?group_id=2888&package_id=40078</a><br>[snip]
<br>> Anyway, as I said, I'd like to use the latest version of gnuplot with Octave
<br>> on Windows. Where do I go from here?
<br><br>The problem with the stock precompiled gnuplot package is that
<br>it does not integrate a patch that I use to generate the version
<br>distributed with octave binary package. This patch allows a real
<br>console mode to gnuplot, which makes it act like on Linux. It
<br>is likely that the stock gnuplot package does not work with octave.
<br><br>The patch is available on octave-forge SVN repository
<br>(admin/Windows/msvc/libs/).
</div>
</div></blockquote>
<p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18639136Re: Octave C++ performance benchmarks?2008-07-24T12:12:48Z2008-07-24T12:12:48ZKyusik ChungThats a good question. Is this help forum for people who are trying
<br>to use Octave and who have not shared all of their code with you
<br>(whether in C++ or octave), or is it just for people who are
<br>developing additions to the Octave code base? You might want to
<br>update the statement on the help page if youd like to keep potential
<br>users like myself from emailing the list.
<br>> The help mailing list exists for the discussion of matters related
<br>> to using and installing Octave.
<br><br>Im of the belief that helping someone out in the forums when they try
<br>to use your software does help the community by creating another great
<br>source of pseudo documentation.
<br><br>In any case, I suppose its a moot point since I am no longer going to
<br>be a user of Octave.
<br><br>Kyusik
<br><br>On Jul 24, 2008, at 11:55 AM, John W. Eaton wrote:
<br><br>> I'm not saying that you can't do what you want to do (assuming that it
<br>> in fact does not violate the terms of the GPL). What I'm asking is
<br>> why our community should bother to help you when what you are doing is
<br>> not helping our community. As I see it, free software is about
<br>> cooperation, not about about us giving and you taking.
<br>>
<br>> jwe
<br><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18639136&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18638697Re: gnuplot zoom2008-07-24T12:03:38Z2008-07-24T12:03:38ZMichael Goffioul-2On Thu, Jul 24, 2008 at 5:59 PM, gOS <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638697&i=0" target="_top" rel="nofollow">bkirklin@...</a>> wrote:
<div class='shrinkable-quote'><br>>
<br>> I'd like to use a version of gnuplot for which the zoom functionality is
<br>> present. I see that it may be possible after looking through the threads on
<br>> this site, though there appeared to be some compiling involved with gcc.
<br>>
<br>> I attempted this but plotting revealed the gnuplot terminal instead of a
<br>> plot. I probably made a mistake somewhere.
<br>>
<br>> I tried this version of gnuplot, extracted: gp423win32.zip
<br>> And I used gnuplot_binary to set the path the pgnuplot.exe
</div><br>I might be wrong, but I thought that mous zooming in gnuplot
<br>was only available in (not yet released) 4.3 version.
<br><br>> Anyhow, I'm using the 3.0.1 (vs 2005) version of Octave for windows located
<br>> here:
<br>> <a href="http://sourceforge.net/project/showfiles.php?group_id=2888&package_id=40078" target="_top" rel="nofollow">http://sourceforge.net/project/showfiles.php?group_id=2888&package_id=40078</a><br>[snip]
<br>> Anyway, as I said, I'd like to use the latest version of gnuplot with Octave
<br>> on Windows. Where do I go from here?
<br><br>The problem with the stock precompiled gnuplot package is that
<br>it does not integrate a patch that I use to generate the version
<br>distributed with octave binary package. This patch allows a real
<br>console mode to gnuplot, which makes it act like on Linux. It
<br>is likely that the stock gnuplot package does not work with octave.
<br><br>The patch is available on octave-forge SVN repository
<br>(admin/Windows/msvc/libs/).
<br><br>Michael.
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638697&i=1" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18638512Re: Octave C++ performance benchmarks?2008-07-24T11:55:33Z2008-07-24T11:55:33ZJohn W. EatonI'm not saying that you can't do what you want to do (assuming that it
<br>in fact does not violate the terms of the GPL). What I'm asking is
<br>why our community should bother to help you when what you are doing is
<br>not helping our community. As I see it, free software is about
<br>cooperation, not about about us giving and you taking.
<br><br>jwe
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638512&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18638341Re: Bug: det() gives false zero2008-07-24T11:46:38Z2008-07-24T11:46:38ZJohn W. EatonOn 24-Jul-2008, Jaroslav Hajek wrote:
<br><br>| Looking at src/DLD-FUNCTIONS/det.cc, it seems that this is a feature -
<br>| if a numerically singular matrix is detected, zero is returned.
<br>| Perhaps John can comment on why he made this choice.
<br><br>If I was the one who did that, then I no longer remember. If it is
<br>the wrong thing to do, then please submit a patch. It should not be
<br>hard to do, as it appears that the Matrix::determinant method already
<br>returns the value you expect in this case.
<br><br>Thanks,
<br><br>jwe
<br>_______________________________________________
<br>Bug-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638341&i=0" target="_top" rel="nofollow">Bug-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/bug-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/bug-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---Bugs-f1896.html" embed="fixTarget[1896]" target="_top" >Octave - Bugs</a></p>tag:www.nabble.com,2006:post-18638506Re: Octave C++ performance benchmarks?2008-07-24T11:43:44Z2008-07-24T11:43:44ZKyusik ChungAccording to the GPL FAQ:
<br><br>> A company is running a modified version of a GPL'ed program on a web
<br>> site. Does the GPL say they must release their modified sources?
<br>> The GPL permits anyone to make a modified version and use it without
<br>> ever distributing it to others. What this company is doing is a
<br>> special case of that. Therefore, the company does not have to
<br>> release the modified sources.
<br><br><br>So what I am talking about is allowed under GPL. If it werent, a lot
<br>of websites out there wouldnt exist (or GPL'd libraries wouldnt be as
<br>widely used). If you disagree with the GPL, you could go with a
<br>different licensing scheme. If you want to stick with GPL, why be
<br>hostile to people who want to use it according to the rules of GPL?
<br><br>In any case, Ive decided that Octave is not a good option for anyone
<br>who is trying to use matrix/vector/lin algebra functionality in their
<br>own C++ code (my original question). As you pointed out, Octave is
<br>for end users to use in the interpreted environment, and using the
<br>libraries in your own code is just a bonus, if it works. Plus, no one
<br>has contributed the time to put together any documentation, nor is
<br>that on any priority list.
<br><br>Boost, FLENS, and O2scl seem like good options for using that kind of
<br>functionality.
<br><br>Kyusik
<br><br>On Jul 24, 2008, at 6:46 AM, John W. Eaton wrote:
<br><div class='shrinkable-quote'><br>> On 22-Jul-2008, Kyusik Chung wrote:
<br>>
<br>> | Sorry for not being more precise. Its not code we are going to
<br>> | distribute. Its back end code that will crunch some data we will
<br>> put
<br>> | on our website, which will be free for consumers to use. I was
<br>> using
<br>> | the term "application" in a very broad sense. Does that qualify?
<br>>
<br>> Qualify in what sense? It does not make the software free for others
<br>> to use. We are sharing Octave with you but you are not sharing your
<br>> work with our community. Why should we go out of our way to help you?
<br>>
<br>> jwe
</div><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638506&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18638310correct behavior of bar(1,1)2008-07-24T11:42:16Z2008-07-24T11:42:16Zde Almeida, Valmor F.<br>Hello,
<br><br>Just wondering whether this is what I should get:
<br><br>octave:1> bar(1,1)
<br>error: invalid row index = 1
<br>error: assignment failed, or no method for `matrix += matrix'
<br>error: evaluating assignment expression near line 126, column 22
<br>error: evaluating if command near line 123, column 3
<br>error: called from `__bar__' in file
<br>`/usr/share/octave/3.0.0/m/plot/__bar__.m'
<br>error: called from `bar' in file `/usr/share/octave/3.0.0/m/plot/bar.m'
<br><br><br>Thanks,
<br><br>--
<br>Valmor
<br><br><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638310&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18638328Re: Errors compiling a C++ class that includes Octave2008-07-24T11:37:33Z2008-07-24T11:37:33ZKyusik ChungJon and Jord, thanks for the tips. Im using a server install of
<br>ubuntu, so apt-cache search octave compile is a good thing to know.
<br>Its too bad that apt-cache search doesnt pick up octave3.0-headers
<br>when you do a search for octave.
<br><br>Kyusik
<br><br>On Jul 22, 2008, at 5:34 PM, Jordi Gutiérrez Hermoso wrote:
<br><div class='shrinkable-quote'><br>> 2008/7/22 Jon Davis <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638328&i=0" target="_top" rel="nofollow">susan.and.jon@...</a>>:
<br>>> Kyusik Chung wrote:
<br>>>> Hi Riza,
<br>>>>
<br>>>> I initially did that, but I don't believe you get the header files
<br>>>> to
<br>>>> be able to compile your own C++ code that includes Octave
<br>>>> libraries. I
<br>>>> was not able to find an "octave-dev" package to install via apt-get
<br>>>> install.
<br>>>
<br>>> What you need is called octave3.0-headers. Using synaptic provides
<br>>> more
<br>>> information than searching with apt-cache, for example.
<br>>
<br>> apt-cache search octave compile
<br>>
<br>> HTH,
<br>> - Jord
</div><br><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18638328&i=1" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18637388Re: Bug: det() gives false zero2008-07-24T10:59:38Z2008-07-24T10:59:38ZJaroslav Hajek-2On Thu, Jul 24, 2008 at 4:15 PM, Charlie Dyson <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18637388&i=0" target="_top" rel="nofollow">charlie@...</a>> wrote:
<div class='shrinkable-quote'><br>> I've been using octave to try to find zeros of the determinants of
<br>> various matrices depending upon some parameter, but have found that
<br>> when the entries in such matrices have wildly different magnitudes
<br>> det() may claim that the matrix is singular when it is not. Using
<br>> Octave 3.0.1 running on XP.
<br>>
<br>> Unfortunately I'll be going on holiday in a few hours so probably
<br>> won't be able to get back to you for a week or so.
<br>>
<br>> Here's some code that reproduces the problem, along with a workaround
<br>> I've been using to cope with it.
<br>>
<br>> Incidentally, I couldn't see any email address at
<br>> <a href="http://www.gnu.org/software/octave/bugs.html" target="_top" rel="nofollow">http://www.gnu.org/software/octave/bugs.html</a>, though I was able to
<br>> find one in the pdf manual.
<br>>
<br>> Many thanks,
<br>>
<br>> Charlie Dyson
<br>>
<br>> -- CUT --
<br>> D = [ -1.001254707117086 0.00125470711708553 1 0 0 0
<br>> 0.00125470711708553 -1.001254707117086 0 1 0 0
<br>> 0 0 -1.050021149616144 -280214911531872 1 0
<br>> 0 0 -5.590061183230268e-021 -0.9499793732213449 0 1
<br>> 1 0 0 0 0 0
<br>> 0 0 0 0 0 1 ];
<br>>
<br>> disp("Compute determinant with det function:")
<br>> disp( det(D) );
<br>> disp("Should not have been zero!");
<br>> disp( "" );
<br>>
<br>> disp("Now try LU decomposition");
<br>> [l,u,p] = lu( D );
<br>> disp("Should all be non-zero:");
<br>> disp( det(l) );
<br>> disp( det(u) );
<br>> disp( det(p) );
<br>> disp("Again we have a false zero");
<br>> disp("");
<br>>
<br>> disp("Now compute det of L, U by product of diagonal");
<br>> detL = 1;
<br>> for n = 1:rows(D)
<br>> detL = detL*l(n,n);
<br>> endfor
<br>> detU = 1;
<br>> for n = 1:rows(D)
<br>> detU = detU*u(n,n);
<br>> endfor
<br>> disp("Det of l,u,p given by:");
<br>> disp( detL );
<br>> disp( detU );
<br>> disp( det(p) );
<br>>
<br>> disp("Correct value of determinant is:");
<br>> disp( detL*detU*det(p) )
<br>> _______________________________________________
<br>> Bug-octave mailing list
<br>> <a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18637388&i=1" target="_top" rel="nofollow">Bug-octave@...</a>
<br>> <a href="https://www.cae.wisc.edu/mailman/listinfo/bug-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/bug-octave</a><br>>
</div><br>Looking at src/DLD-FUNCTIONS/det.cc, it seems that this is a feature -
<br>if a numerically singular matrix is detected, zero is returned.
<br>Perhaps John can comment on why he made this choice.
<br><br>--
<br>RNDr. Jaroslav Hajek
<br>computing expert
<br>Aeronautical Research and Test Institute (VZLU)
<br>Prague, Czech Republic
<br>url: www.highegg.matfyz.cz
<br>_______________________________________________
<br>Bug-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18637388&i=2" target="_top" rel="nofollow">Bug-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/bug-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/bug-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---Bugs-f1896.html" embed="fixTarget[1896]" target="_top" >Octave - Bugs</a></p>tag:www.nabble.com,2006:post-18636515Re: Behavior of 'mkdir'2008-07-24T10:08:38Z2008-07-24T10:08:38ZJohn W. EatonOn 24-Jul-2008, WMennerich wrote:
<br><br>| But another thing which could increase the Octave/Matlab compatiblility:
<br>|
<br>| The message string of the mkdir-function differs in the discussed case:
<br>|
<br>| Matlab wrotes: 'Directory already exists.' (also with the point at the end)
<br>|
<br>| Octave wrotes: 'File exists'
<br>|
<br>| I would suggest to fit the message string given back by the octave mkdir to
<br>| the output which comes from matlab to make strcmp-constructs more compatible
<br>| between Matlab and Octave. Also because 'File exists' may be
<br>| misleading: Sure, in UnixU/Linux everything is a 'file', also directories.
<br>| But since
<br>| the name of this function is 'mkdir' and not 'mkfile' or something else, the
<br>| message should have a relation to 'directory'.
<br><br>I have no desire to change error messages so that they match exactly
<br>the Matlab error messages (you would do that, but not change the
<br>behavior of the mkdir function?).
<br><br>In Octave, this message comes from the C-level strerror function.
<br>The text is system dependent. If Octave is modified to support
<br>different locale settings, then the message will change if you change
<br>your locale setting to another language. So I think checking the
<br>text to decide what to do is probably not a good programming strategy.
<br><br>I think it might be best to simply change the behavior to match
<br>Matlab's in this case. However, I don't think that we need to
<br>duplicate the behavior exactly. I see that Matlab also returns
<br>status=1 and msg="Directory already exists." when the "directory" is
<br>just a regular file. That is probably not a good thing to do. How
<br>about the attached patch?
<br><br>jwe
<br><br><br /># HG changeset patch
<br># User John W. Eaton <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18636515&i=0" target="_top" rel="nofollow">jwe@...</a>>
<br># Date 1216919016 14400
<br># Node ID b6d4c644b4b61ada8f5f7ede36072a9b64005389
<br># Parent 4f9e8eeb2059cf6d20e821958233d4e773ef4cff
<br>Fmkdir: improve compatibility
<br><br>diff --git a/src/ChangeLog b/src/ChangeLog
<br>--- a/src/ChangeLog
<br>+++ b/src/ChangeLog
<br>@@ -1,3 +1,8 @@
<br>+2008-07-24 John W. Eaton <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18636515&i=1" target="_top" rel="nofollow">jwe@...</a>>
<br>+
<br>+ * dirfns.cc (Fmkdir): If directory already exists, return status =
<br>+ true, but also set error message.
<br>+
<br> 2008-07-23 John W. Eaton <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18636515&i=2" target="_top" rel="nofollow">jwe@...</a>>
<br>
<br> * ov-usr_fcn.cc (octave_user_function::do_multi_index_op):
<br>diff --git a/src/dirfns.cc b/src/dirfns.cc
<br>--- a/src/dirfns.cc
<br>+++ b/src/dirfns.cc
<br>@@ -257,16 +257,31 @@
<br> {
<br> std::string msg;
<br>
<br>- int status = file_ops::mkdir (file_ops::tilde_expand (dirname),
<br>- 0777, msg);
<br>+ dirname = file_ops::tilde_expand (dirname);
<br>
<br>- if (status < 0)
<br>+ file_stat fs (dirname);
<br>+
<br>+ if (fs && fs.is_dir ())
<br> {
<br>+ // For compatibility with Matlab, we return true when the
<br>+ // directory already exists.
<br>+
<br> retval(2) = "mkdir";
<br>- retval(1) = msg;
<br>+ retval(1) = "directory exists";
<br>+ retval(0) = true;
<br> }
<br> else
<br>- retval(0) = true;
<br>+ {
<br>+ int status = file_ops::mkdir (dirname, 0777, msg);
<br>+
<br>+ if (status < 0)
<br>+ {
<br>+ retval(2) = "mkdir";
<br>+ retval(1) = msg;
<br>+ }
<br>+ else
<br>+ retval(0) = true;
<br>+ }
<br> }
<br> else
<br> print_usage ();
<br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18636295Re: Behavior of 'mkdir'2008-07-24T09:57:11Z2008-07-24T09:57:11ZShai Ayal-2<div dir="ltr"><br><br><div class="gmail_quote">On Thu, Jul 24, 2008 at 7:31 PM, WMennerich <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18636295&i=0" target="_top" rel="nofollow">W.Mennerich@...</a>> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
First:<br>
<br>
This (simple) solution should work in both environments:<br>
<br>
><br>
>[ack msg]=mkdir('testDirectory')<br>
><br>
>if ~ack<br>
> if strcmp(msg,'File exists')<br>
> % 'Wrong alert' by octave, -->do nothing<br>
> else<br>
> % some other thing is wrong, e.g. throw an errror here.<br>
> end<br>
>end</blockquote><div><br>Maybe you can submit a patch like this to the mkdir documentation?<br> <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>
But another thing which could increase the Octave/Matlab compatiblility:<br>
<br>
The message string of the mkdir-function differs in the discussed case:<br>
<br>
Matlab wrotes: 'Directory already exists.' (also with the point at the end)<br>
<br>
Octave wrotes: 'File exists'<br>
<br>
<br>
I would suggest to fit the message string given back by the octave mkdir to<br>
the output which comes from matlab to make strcmp-constructs more compatible<br>
between Matlab and Octave. Also because 'File exists' may be<br>
misleading: Sure, in UnixU/Linux everything is a 'file', also directories.<br>
But since<br>
the name of this function is 'mkdir' and not 'mkfile' or something else, the<br>
message should have a relation to 'directory'</blockquote><div><br>So can you submit a patch?<br><br> Shai</div></div><br></div>
<p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18636044Re: Starting m-file from command line/shortcut in Windows2008-07-24T09:50:54Z2008-07-24T09:50:54ZgOSWhy not add the m-file to your .octaverc file? Or even at startup of octave ask if you wish to run the m file and then do it?
<br><br>if(isequal(input('run mfile?','s'),'y')), mfile, end;
<br><br>Or you could make an app generate a new .octaverc, copy the original to a safe place, and then after octave terminates reinstate the original .octaverc. Probably could do it with a bat file or some other script.
<br><br>Bat file would look something like:
<br><br># copy original .octaverc to some other directory or rename to .octaverc.bak
<br># create new .octaverc
<br># C:/...pathtooctave.../octave/bin/octave
<br># replace .octaverc
<br><br>.octaverc would look something like:
<br><br>more('off')
<br>mFileToRun(inputs)
<br>exit<p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18636032Re: Matlab compliance2008-07-24T09:46:23Z2008-07-24T09:46:23ZJohn W. EatonOn 24-Jul-2008, Michael D. Godfrey wrote:
<br><br>| >
<br>| > We used to have built-in variables that controlled things like that,
<br>| > but it was the wrong thing to do for various reasons.
<br>| This seems still true now.
<br><br>It is true for variables the control the behavior at run-time only,
<br>but if we introduced a directive that takes effect when a file is
<br>parsed, and that is limited to the scope of the file or enclosing
<br>function, then I think we could avoid most (or all) of the problems.
<br>To be sure probably requires a little more thought and some discussion
<br>about the design, but I think it is possible. Still, someone has to
<br>do the work to implement, and at the moment I don't have the time or
<br>interest to do it myself.
<br><br>jwe
<br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18635652Re: Behavior of 'mkdir'2008-07-24T09:31:54Z2008-07-24T09:31:54ZWMennerichFirst:
<br><br>This (simple) solution should work in both environments:
<br><br>>
<br>>[ack msg]=mkdir('testDirectory')
<br>>
<br>>if ~ack
<br>> if strcmp(msg,'File exists')
<br>> % 'Wrong alert' by octave, -->do nothing
<br>> else
<br>> % some other thing is wrong, e.g. throw an errror here.
<br>> end
<br>>end
<br><br>But another thing which could increase the Octave/Matlab compatiblility:
<br><br>The message string of the mkdir-function differs in the discussed case:
<br><br>Matlab wrotes: 'Directory already exists.' (also with the point at the end)
<br><br>Octave wrotes: 'File exists'
<br><br><br>I would suggest to fit the message string given back by the octave mkdir to
<br>the output which comes from matlab to make strcmp-constructs more compatible
<br>between Matlab and Octave. Also because 'File exists' may be
<br>misleading: Sure, in UnixU/Linux everything is a 'file', also directories. But since
<br>the name of this function is 'mkdir' and not 'mkfile' or something else, the
<br>message should have a relation to 'directory'.
<br><br><br>Second:
<br><br>Yes, you may be right, one can not conclude that programmers using 'mkdir' have
<br>more background than others: This conclusion came from my own experience: With my first matlab programms, I was happy that it runs somehow, but I was far away from playing
<br>with directories and stuff like that, e.g user inputs, loading and writing data, ....
<br>;-)
<br><br>Wolfgang<p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18639219ps files read under octave2008-07-24T09:09:19Z2008-07-24T09:09:19ZCecile LimborgHello -
<br><br>Congratulations for all the tools developed so far.
<br>I can now use Octave for most of my data processing.
<br><br>I have a question related to loading image files.
<br>I converted a gif file into ps, since I believed that octave could load
<br>it . My ps file is correctly written, and reads well under ghostview.
<br>However it does not read with octave either with loadimage or image_viewer.
<br><br>Do you have any recommendation ?
<br><br>Thank you
<br>Cecile,
<br>Accelerator Physicist
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18639219&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18635402Starting m-file from command line/shortcut in Windows2008-07-24T09:05:50Z2008-07-24T09:05:50ZAllen.Windhorn-2Hello all,
<br><br>I have Octave 3.0.0 on Windows XP. I would like to start an Octave
<br>m-file from the command line and also from a desktop shortcut, have it
<br>run until the user is done, then automatically close Octave (the last
<br>part is optional).
<br><br>I used the following command from a command shell:
<br><br>H:\>"c:\Program Files\Octave\bin\octave-3.0.0" -q -persist
<br>"h:/work/matlab/unbal.m"
<br><br>Octave starts and brings up the program but dies:
<br>--------
<br> - Use `pkg list' to see a list of installed packages.
<br> - SciTE editor installed. Use `edit' to start the editor.
<br> - MSYS shell available (c:\Program Files\Octave\msys).
<br> - Graphics backend: jhandles.
<br><br><br> CALCULATE VOLTAGE UNBALANCE
<br><br>Enter phase currents. Note: all currents must be > 0 for case 3
<br>Enter A phase current p.u. <0>: panic: Segmentation violation --
<br>stopping myself...
<br>attempting to save variables to `octave-core'...
<br>save to `octave-core' complete
<br>---------
<br><br>When I use the same command in a shortcut I either get a blank screen or
<br>Octave starts and hangs without giving me a prompt. Without the options
<br>the result is the same.
<br><br>Is this a bug, or am I doing something wrong? Do I need a different
<br>input method in batch mode?
<br><br>Following is a bit of the m-file code:
<br><br>---------
<br>%unbal -- Calculate voltage unbalance given winding currents
<br>% Using assumptions about distribution of currents, and given three
<br>% winding currents, power factor, and reactances, calculate phase
<br>% voltages and voltage unbalance.
<br>%
<br>% Assumptions differ by case:
<br>%
<br>% 1. Input currents as phasors (power factor ignored)
<br>%
<br>% 2. Assume all currents spaced at 120 degrees, line to neutral
<br>%
<br>% 3. Assume no zero sequence current
<br>%
<br>% Allen Windhorn, 24 May 2000
<br>%
<br>% Initialize things:
<br>%
<br>format short g; %Display results to more precision
<br>clear all
<br>%
<br>a = -0.5+j*(sqrt(3)/2);
<br>r2d = 180/pi;
<br>TRUE = (1==1);
<br>FALSE = (1==2);
<br>%
<br>% Print header
<br>%
<br>disp(' ');
<br>disp(' CALCULATE VOLTAGE UNBALANCE');
<br>disp(' ')
<br>disp('Enter phase currents. Note: all currents must be > 0 for case
<br>3');
<br>%
<br>% Get values of phase currents, power factor, reactances
<br>% Perform error checking to make sure values entered
<br>%
<br>Ia = input('Enter A phase current p.u. <0>: ');
<br>if isempty(Ia)
<br> Ia = 0;
<br>end
<br>%
<br>Ib = input('Enter B phase current p.u. <0>: ');
<br>if isempty(Ib)
<br> Ib = 0;
<br>-----------
<br><br>Thanks in advance for any advice.
<br><br>Regards,
<br>Allen
<br>--
<br>Allen Windhorn, P.E. (MN) (507) 345-2782
<br>Kato Engineering
<br>P.O. Box 8447, N. Mankato, MN 56002
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18635402&i=0" target="_top" rel="nofollow">Allen.Windhorn@...</a>
<br><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18635402&i=1" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18639086siso_ldpc_fast_mex.mex is not a valid shared library!2008-07-24T09:03:48Z2008-07-24T09:03:48ZHonore Tapamo<br>Dear Octave users/developers,
<br><br>I am getting the following error when calling a mex function within octave
<br><br> "....siso_ldpc_fast_mex.mex is not a valid shared library".
<br><br>please could someone tell me what the problem here is. Is it an issue with
<br>static/shared libraries.
<br><br>For information, I used static libraries to build octave.
<br><br>Thank you very much for your help.
<br><br>Regards,
<br>Honore.
<br><br><br><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18639086&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18634994gnuplot zoom2008-07-24T08:59:25Z2008-07-24T08:59:25ZgOSI'd like to use a version of gnuplot for which the zoom functionality is present. I see that it may be possible after looking through the threads on this site, though there appeared to be some compiling involved with gcc.
<br><br>I attempted this but plotting revealed the gnuplot terminal instead of a plot. I probably made a mistake somewhere.
<br><br>I tried this version of gnuplot, extracted: gp423win32.zip
<br>And I used gnuplot_binary to set the path the pgnuplot.exe
<br><br>Anyhow, I'm using the 3.0.1 (vs 2005) version of Octave for windows located here: <a href="http://sourceforge.net/project/showfiles.php?group_id=2888&package_id=40078" target="_top" rel="nofollow">http://sourceforge.net/project/showfiles.php?group_id=2888&package_id=40078</a><br><br>Jhandles is of course an option for zooming, but on one machine Octave would crash (Java was updated past 1.5.0). Another problem with Jhandles is that the title portion would not allow for more than 4 lines of text, and the default for most plots that I make is 5 lines (due to versioning information). There were other factors that went into deciding not to use Jhandles for the moment, but I've since forgotten most of them. I just remember we were going to try it out again when a new version was released.
<br><br>Anyway, as I said, I'd like to use the latest version of gnuplot with Octave on Windows. Where do I go from here?<p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18634946Re: do exp(-0.5*z.*z) in .oct file?2008-07-24T08:56:10Z2008-07-24T08:56:10ZMichael Creel<br><blockquote class="quote light-black dark-border-color"><div class="quote light-border-color">
<div class="quote-author" style="font-weight: bold;">Michael Creel wrote:</div>
<div class="quote-message shrinkable-quote">Hello all,
<br>I'm trying to write the C++ code for an .oct that will perform the
<br>equivalent of exp(-0.5*z.*z), where z is a vector or matrix. I think
<br>that Matrix.map will let me do the exp() to all elements, but I can't
<br>figure out how to do the element-by-element squaring.
<br>Thanks, Michael
<br>_______________________________________________
<br>Help-octave mailing list
<br>Help-octave@octave.org
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a></div>
</div></blockquote>
found it: product(z,z) does the trick. M<p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18633282Re: Behavior of 'mkdir'2008-07-24T07:34:31Z2008-07-24T07:34:31ZJohn W. EatonOn 24-Jul-2008, WMennerich wrote:
<br><br>| I already found a way to implement my code for working fine with both,
<br>| Matlab and Octave.
<br><br>How? What is your solution?
<br><br>| The more I thing about it, the more I find that octave does
<br>| it in the best way now.
<br><br>Not if what you want is for Matlab code to work without change in
<br>Octave.
<br><br>| Even if it makes trouble to run matlab written code with octave,
<br>| the people which use the 'mkdir' command normally are not new
<br>| to Matlab/Ocatave and may have enough background to come clear with
<br>| these differences.
<br><br>I don't think this conclusion is justified. How do you know who is
<br>using the mkdir function or what their level of expertise is?
<br><br>jwe
<br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18635908Re: Matlab compliance2008-07-24T07:33:13Z2008-07-24T07:33:13ZMichael D. Godfrey>
<br>> We used to have built-in variables that controlled things like that,
<br>> but it was the wrong thing to do for various reasons.
<br>This seems still true now.
<br><br>Thanks for your reply.
<br><br>Michael
<br><br><br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18633001Bug: det() gives false zero2008-07-24T07:15:40Z2008-07-24T07:15:40ZCharlie DysonI've been using octave to try to find zeros of the determinants of
<br>various matrices depending upon some parameter, but have found that
<br>when the entries in such matrices have wildly different magnitudes
<br>det() may claim that the matrix is singular when it is not. Using
<br>Octave 3.0.1 running on XP.
<br><br>Unfortunately I'll be going on holiday in a few hours so probably
<br>won't be able to get back to you for a week or so.
<br><br>Here's some code that reproduces the problem, along with a workaround
<br>I've been using to cope with it.
<br><br>Incidentally, I couldn't see any email address at
<br><a href="http://www.gnu.org/software/octave/bugs.html" target="_top" rel="nofollow">http://www.gnu.org/software/octave/bugs.html</a>, though I was able to
<br>find one in the pdf manual.
<br><br>Many thanks,
<br><br>Charlie Dyson
<br><br>-- CUT --
<br>D = [ -1.001254707117086 0.00125470711708553 1 0 0 0
<br> 0.00125470711708553 -1.001254707117086 0 1 0 0
<br> 0 0 -1.050021149616144 -280214911531872 1 0
<br> 0 0 -5.590061183230268e-021 -0.9499793732213449 0 1
<br> 1 0 0 0 0 0
<br> 0 0 0 0 0 1 ];
<br><br>disp("Compute determinant with det function:")
<br>disp( det(D) );
<br>disp("Should not have been zero!");
<br>disp( "" );
<br><br>disp("Now try LU decomposition");
<br>[l,u,p] = lu( D );
<br>disp("Should all be non-zero:");
<br>disp( det(l) );
<br>disp( det(u) );
<br>disp( det(p) );
<br>disp("Again we have a false zero");
<br>disp("");
<br><br>disp("Now compute det of L, U by product of diagonal");
<br>detL = 1;
<br>for n = 1:rows(D)
<br> detL = detL*l(n,n);
<br>endfor
<br>detU = 1;
<br>for n = 1:rows(D)
<br> detU = detU*u(n,n);
<br>endfor
<br>disp("Det of l,u,p given by:");
<br>disp( detL );
<br>disp( detU );
<br>disp( det(p) );
<br><br>disp("Correct value of determinant is:");
<br>disp( detL*detU*det(p) )
<br>_______________________________________________
<br>Bug-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18633001&i=0" target="_top" rel="nofollow">Bug-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/bug-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/bug-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---Bugs-f1896.html" embed="fixTarget[1896]" target="_top" >Octave - Bugs</a></p>tag:www.nabble.com,2006:post-18632273Re: Behavior of 'mkdir'2008-07-24T06:46:36Z2008-07-24T06:46:36ZWMennerich<br><blockquote class="quote light-black dark-border-color"><div class="quote light-border-color">
<div class="quote-author" style="font-weight: bold;">Jaroslav Hajek-2 wrote:</div>
<div class="quote-message">Wolfgang, do you want to fix this yourself?
</div>
</div></blockquote>
I already found a way to implement my code for working fine with both,
<br>Matlab and Octave. So, there is no need for me to change this.
<br>The more I thing about it, the more I find that octave does
<br>it in the best way now.
<br>Even if it makes trouble to run matlab written code with octave,
<br>the people which use the 'mkdir' command normally are not new
<br>to Matlab/Ocatave and may have enough background to come clear with
<br>these differences.
<br><br>Best, Wolfgang<p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18632761Re: Octave C++ performance benchmarks?2008-07-24T06:46:23Z2008-07-24T06:46:23ZJohn W. EatonOn 22-Jul-2008, Kyusik Chung wrote:
<br><br>| Sorry for not being more precise. Its not code we are going to
<br>| distribute. Its back end code that will crunch some data we will put
<br>| on our website, which will be free for consumers to use. I was using
<br>| the term "application" in a very broad sense. Does that qualify?
<br><br>Qualify in what sense? It does not make the software free for others
<br>to use. We are sharing Octave with you but you are not sharing your
<br>work with our community. Why should we go out of our way to help you?
<br><br>jwe
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18632761&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18632481Re: datestr bug2008-07-24T06:24:32Z2008-07-24T06:24:32ZJohn W. EatonOn 23-Jul-2008, Ben Abbott wrote:
<br><br>| Another possible solution (hack) is to replace the offending line with
<br>|
<br>| nd = strfind (str, "d");
<br>| nd = setdiff (nd, strfind (str,"day"));
<br>| nd = setdiff (nd, strfind (str,"dnesday"));
<br>| str(nd) = names_d{weekday (datenum (v(i,1), v(i,2), v(i,3)))};
<br><br>I'm not applying this change because I think we should be able to come
<br>up with a proper solution that is not just a band-aid kluge fix.
<br><br>| In either event, an additional test should be added.
<br>|
<br>| %!assert(datestr(testtime+[0 0 3 0 0 0],"dddd"),"Wednesday")
<br>|
<br>| Comments?
<br><br>I added this as an "xtest" that is expected to fail.
<br><br>Thanks,
<br><br>jwe
<br>_______________________________________________
<br>Bug-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18632481&i=0" target="_top" rel="nofollow">Bug-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/bug-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/bug-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---Bugs-f1896.html" embed="fixTarget[1896]" target="_top" >Octave - Bugs</a></p>tag:www.nabble.com,2006:post-18632462Matlab compliance2008-07-24T06:13:40Z2008-07-24T06:13:40ZJohn W. EatonOn 23-Jul-2008, Michael D. Godfrey wrote:
<br><br>| Have you thought about a "Matlab_mode" setting
<br>| so that users could run in Matlab compliant mode or
<br>| a mode that is accepted as better practice? This could
<br>| be a vector of settings to allow "picking and choosing."
<br>|
<br>| You may have already decided about this. But, since
<br>| I have not seen it mentioned, I am mentioning it now
<br>| in the hope that it will get some review and a decision.
<br><br>We used to have built-in variables that controlled things like that,
<br>but it was the wrong thing to do for various reasons. Two of the
<br>biggest problems were that they could only affect behavior at
<br>run-time, not when the function was parsed, and they were global
<br>settings.
<br><br>It might be possible to have parse-time directives to do similar
<br>things, provided that the directive only applies to the current
<br>function (or perhaps file) scope. It might be worth doing this, but I
<br>don't have the time at the moment to work on it.
<br><br>jwe
<br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18632392Re: Octave 3.1.51 available for ftp2008-07-24T06:04:10Z2008-07-24T06:04:10ZJohn W. EatonOn 23-Jul-2008, Thomas Treichl wrote:
<br><br>| There is one thing that I still don't understand when I compile the sources and
<br>| maybe someone can give me a hand here: If I try to link liboctave then I get the
<br>| following error (I know this since building 2.9.12 but I always thought that
<br>| this may be a special problem of my machine?!):
<br>|
<br>| ld: Undefined symbols:
<br>| _fftw_destroy_plan referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3.3.dylib
<br>| _fftw_execute_dft referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3.3.dylib
<br>| _fftw_execute_dft_r2c referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3.3.dylib
<br>| _fftw_import_system_wisdom referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3.3.dylib
<br>| _fftw_plan_many_dft referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3.3.dylib
<br>| _fftw_plan_many_dft_r2c referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3.3.dylib
<br>| _fftwf_destroy_plan referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3f.3.dylib
<br>| _fftwf_execute_dft referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3f.3.dylib
<br>| _fftwf_execute_dft_r2c referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3f.3.dylib
<br>| _fftwf_import_system_wisdom referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3f.3.dylib
<br>| _fftwf_plan_many_dft referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3f.3.dylib
<br>| _fftwf_plan_many_dft_r2c referenced from liboctave expected to be defined in
<br>| /tmp/dependencies-i386/lib/libfftw3f.3.dylib
<br>| /usr/bin/libtool: internal link edit command failed
<br>| make[2]: *** [liboctinterp.dylib] Error 1
<br>| make[1]: *** [src] Error 2
<br>| make: *** [all] Error 2
<br>| solvedeps-3.1.51.sh: Building Octave.app has failed !!
<br>|
<br>| My workaround for this is to apply a patch of the following form:
<br>|
<br>| --- octave-3.1.51.orig/src/Makefile.in
<br>| +++ octave-3.1.51/src/Makefile.in
<br>|
<br>| OCTINTERP_LINK_DEPS = \
<br>| -L../liboctave $(LIBOCTAVE) -L../libcruft $(LIBCRUFT) $(LIBS) $(FLIBS) \
<br>| - $(OPENGL_LIBS)
<br>| + $(OPENGL_LIBS) $(FFTW_LIBS)
<br>|
<br>| OCT_LINK_DEPS = \
<br>| -L../libcruft $(LIBCRUFT) -L../liboctave $(LIBOCTAVE) \
<br>|
<br>| then I can link liboctave.dylib without any problems left. Any ideas what may
<br>| cause the above problem?
<br><br>As far as I can tell, fftw functions are only referenced directly in
<br>liboctave, so I don't see why you should need to also link the fftw
<br>library directly with liboctinterp. If this is needed for fftw, why
<br>is it not also needed for the suitesparse libraries?
<br><br>jwe
<br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18632345do exp(-0.5*z.*z) in .oct file?2008-07-24T06:01:54Z2008-07-24T06:01:54ZMichael CreelHello all,
<br>I'm trying to write the C++ code for an .oct that will perform the
<br>equivalent of exp(-0.5*z.*z), where z is a vector or matrix. I think
<br>that Matrix.map will let me do the exp() to all elements, but I can't
<br>figure out how to do the element-by-element squaring.
<br>Thanks, Michael
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18632345&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18632145Re: make error about log2 for octave-3.1.51 (Not 3.1.50)2008-07-24T05:59:59Z2008-07-24T05:59:59ZJaroslav Hajek-2On Thu, Jul 24, 2008 at 11:54 AM, Tatsuro MATSUOKA
<br><<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18632145&i=0" target="_top" rel="nofollow">tmacchant@...</a>> wrote:
<div class='shrinkable-quote'><br>> Hello Jaroslav
<br>>
<br>> Thank you for your reply
<br>>
<br>>
<br>>> This looks really weird given that there is no / token on the line. I
<br>>> suspect that
<br>>> you have log2 defined to some macro and that causes problems.
<br>>
<br>> I do not define log2 myself. Perhaps gcc on cygwin defines it.
<br>> Since octave 3.0.1, the log2 fuction has no problem at building.
<br>>
</div><br>Yes, that's what I meant. I didn't mean to accuse you of causing the mess :)
<br><div class='shrinkable-quote'><br>>
<br>>
<br>>> Can you try running:
<br>>>
<br>>> /opt/octave/gcc-3.4.4dw2/bin/g++ -E -C
<br>>> -I/opt/octave/gcc-3.4.4dw2/include -I/usr/include
<br>>> -I/opt/octave/octave-3.1.51/include -I. -I.. -I../liboctave -I../src
<br>>> -I../libcruft/misc
<br>>> -DHAVE_CONFIG_H -mieee-fp -Wall -W -Wshadow -Wold-style-cast -g -O2 Cell.cc
<br>>>
<br>>> in the directory where the make failure occured and posting the output?
<br>> I will show the results of which the log2 related.
</div><br>Sorry, but this tells me nothing new. I wanted to see the lines that
<br>correspond to the
<br>log2 methods definitions in ov.h and ov-base.h. They will have
<br>different line numbers in this output - that's why I wanted the whole
<br>output. If you don't want to clutter the ML with large dumps, please
<br>use an attachment or some service like pastebin.
<br>I suspect that some macro like
<br><br>#define log2(x) (log(x)/log(2))
<br><br>or similar interferes with the definition
<br><br>virtual octave_value log2 (void) const;
<br><br>turning it into a nonsense. As John says, it will probably be
<br>necessary to add #undef log2 to some proper place. Anyway, he is right
<br>- please report this to <a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18632145&i=1" target="_top" rel="nofollow">bug@...</a>, as this list is not dedicated
<br>to bug fixing.
<br><div class='shrinkable-quote'><br>>
<br>> 493: /* Define to 1 if you have the `log2' function. */
<br>> 494: /* #undef HAVE_LOG2 */
<br>>
<br>>
<br>> The below lines perhaps are not related but I will show them as a reference
<br>> 47686: extern double xlog2 (double x);
<br>> 47687: extern Complex xlog2 (const Complex& x);
<br>> 47688: extern double xlog2 (double x, int& exp);
<br>> 47689: extern Complex xlog2 (const Complex& x, int& exp);
<br>>
<br>> 47740: extern float xlog2 (float x);
<br>> 47741: extern FloatComplex xlog2 (const FloatComplex& x);
<br>> 47742: extern float xlog2 (float x, int& exp);
<br>> 47743: extern FloatComplex xlog2 (const FloatComplex& x, int& exp);
<br>>
<br>> To see the how configure detect log2
<br>> configure:25965: checking for log2
<br>> configure:26029: /opt/octave/gcc-3.4.4dw2/bin/gcc -o conftest.exe -g -O2
<br>> -I/opt/octave/gcc-3.4.4dw2/include -I/usr/include -I/opt/octave/octave-3.1.51/include
<br>> -L/opt/octave/gcc-3.4.4dw2/lib -L/opt/octave/octave-3.1.51/lib conftest.c -ldl -lwsock32 -lblas -lhdf5
<br>> -lz -lm -lwsock32 >&5
<br>> conftest.c:265: warning: conflicting types for built-in function 'log2'
<br>> /cygdrive/c/DOCUME~1/Tatsu/LOCALS~1/Temp/cccEhDNQ.o: In function `main':
<br>> /home/octaves/octave-3.1.51/conftest.c:284: undefined reference to `_log2'
<br>> collect2: ld returned 1 exit status
<br>> configure:26035: $? = 1
<br>> configure: failed program was:
<br>> | /* confdefs.h. */
<br>> | #define PACKAGE_NAME ""
<br>> | #define PACKAGE_TARNAME ""
<br>> | #define PACKAGE_VERSION ""
<br>> | #define PACKAGE_STRING ""
<br>> | #define PACKAGE_BUGREPORT ""
<br>> | #define OCTAVE_SOURCE 1
<br>> | #define _GNU_SOURCE 1
<br>> | #define STDC_HEADERS 1
<br>> | #define HAVE_SYS_TYPES_H 1
<br>> | #define HAVE_SYS_STAT_H 1
<br>> | #define HAVE_STDLIB_H 1
<br>> | #define HAVE_STRING_H 1
<br>> | #define HAVE_MEMORY_H 1
<br>> | #define HAVE_STRINGS_H 1
<br>> | #define HAVE_INTTYPES_H 1
<br>> | #define HAVE_STDINT_H 1
<br>> | #define HAVE_UNISTD_H 1
<br>> | #define SEPCHAR ':'
<br>> | #define SEPCHAR_STR ":"
<br>> | #define __NO_MATH_INLINES 1
<br>> | #define CXX_NEW_FRIEND_TEMPLATE_DECL 1
<br>> | #define CXX_ISO_COMPLIANT_LIBRARY 1
<br>> | #define CXX_BROKEN_REINTERPRET_CAST 1
<br>> | #define CXX_ABI gnu_v3
<br>> | #define HAVE_LIBM 1
<br>> | #define HAVE_QHULL 1
<br>> | #define HAVE_PCRE 1
<br>> | #define HAVE_REGEXEC 1
<br>> | #define HAVE_REGEX 1
<br>> | #define HAVE_ZLIB_H 1
<br>> | #define HAVE_ZLIB 1
<br>> | #define HAVE_HDF5_H 1
<br>> | #define HAVE_HDF5 1
<br>> | #define HAVE_H5GGET_NUM_OBJS 1
<br>> | #define HAVE_FFTW3 1
<br>> | #define HAVE_GLPK_H 1
<br>> | #define HAVE_GLPK 1
<br>> | #define HAVE_CURL_CURL_H 1
<br>> | #define HAVE_CURL 1
<br>> | #define HAVE_MAGICK 1
<br>> | #define HAVE_GL_GL_H 1
<br>> | #define HAVE_GL_GLU_H 1
<br>> | #define HAVE_OPENGL 1
<br>> | #define HAVE_FLTK 1
<br>> | #define HAVE_IEEE754_DATA_FORMAT 1
<br>> | #define F77_FUNC(name,NAME) name ## _
<br>> | #define F77_FUNC_(name,NAME) name ## __
<br>> | #define HAVE_BLAS 1
<br>> | #define HAVE_SUITESPARSE_AMD_H 1
<br>> | #define HAVE_AMD 1
<br>> | #define HAVE_SUITESPARSE_UMFPACK_H 1
<br>> | #define HAVE_UMFPACK 1
<br>> | #define UMFPACK_SEPARATE_SPLIT 1
<br>> | #define HAVE_SUITESPARSE_COLAMD_H 1
<br>> | #define HAVE_COLAMD 1
<br>> | #define HAVE_SUITESPARSE_CCOLAMD_H 1
<br>> | #define HAVE_CCOLAMD 1
<br>> | #define HAVE_SUITESPARSE_CHOLMOD_H 1
<br>> | #define HAVE_CHOLMOD 1
<br>> | #define HAVE_SUITESPARSE_CS_H 1
<br>> | #define HAVE_CXSPARSE 1
<br>> | #define HAVE_GETHOSTNAME 1
<br>> | #define HAVE_GETPWNAM 1
<br>> | #define HAVE_LIBWSOCK32 1
<br>> | #define HAVE_DEV_T 1
<br>> | #define HAVE_INO_T 1
<br>> | #define HAVE_NLINK_T 1
<br>> | #define HAVE_NLINK_T 1
<br>> | #define HAVE_LONG_LONG_INT 1
<br>> | #define HAVE_UNSIGNED_LONG_LONG_INT 1
<br>> | #define HAVE_SIGSET_T 1
<br>> | #define HAVE_SIG_ATOMIC_T 1
<br>> | #define SIZEOF_SHORT 2
<br>> | #define SIZEOF_INT 4
<br>> | #define SIZEOF_LONG 4
<br>> | #define SIZEOF_LONG_LONG 8
<br>> | #define HAVE_ALLOCA_H 1
<br>> | #define HAVE_ALLOCA 1
<br>> | #define NPOS std::string::npos
<br>> | #define HAVE_PLACEMENT_DELETE 1
<br>> | #define HAVE_DYNAMIC_AUTO_ARRAYS 1
<br>> | #define STDC_HEADERS 1
<br>> | #define HAVE_DIRENT_H 1
<br>> | #define TIME_WITH_SYS_TIME 1
<br>> | #define HAVE_SYS_WAIT_H 1
<br>> | #define HAVE_ASSERT_H 1
<br>> | #define HAVE_CURSES_H 1
<br>> | #define HAVE_DLFCN_H 1
<br>> | #define HAVE_FCNTL_H 1
<br>> | #define HAVE_FLOAT_H 1
<br>> | #define HAVE_GRP_H 1
<br>> | #define HAVE_IEEEFP_H 1
<br>> | #define HAVE_INTTYPES_H 1
<br>> | #define HAVE_LIMITS_H 1
<br>> | #define HAVE_LOCALE_H 1
<br>> | #define HAVE_MEMORY_H 1
<br>> | #define HAVE_NCURSES_H 1
<br>> | #define HAVE_POLL_H 1
<br>> | #define HAVE_PTHREAD_H 1
<br>> | #define HAVE_PWD_H 1
<br>> | #define HAVE_STDINT_H 1
<br>> | #define HAVE_STDLIB_H 1
<br>> | #define HAVE_STRING_H 1
<br>> | #define HAVE_SYS_IOCTL_H 1
<br>> | #define HAVE_SYS_PARAM_H 1
<br>> | #define HAVE_SYS_POLL_H 1
<br>> | #define HAVE_SYS_RESOURCE_H 1
<br>> | #define HAVE_SYS_SELECT_H 1
<br>> | #define HAVE_SYS_STAT_H 1
<br>> | #define HAVE_SYS_TIME_H 1
<br>> | #define HAVE_SYS_TIMES_H 1
<br>> | #define HAVE_SYS_TYPES_H 1
<br>> | #define HAVE_SYS_UTSNAME_H 1
<br>> | #define HAVE_SYS_UTIME_H 1
<br>> | #define HAVE_TERMCAP_H 1
<br>> | #define HAVE_UNISTD_H 1
<br>> | #define HAVE_UTIME_H 1
<br>> | #define HAVE_SSTREAM 1
<br>> | #define HAVE_TERMIO_H 1
<br>> | #define HAVE_GLOB_H 1
<br>> | #define HAVE_FNMATCH_H 1
<br>> | #define HAVE_FNMATCH 1
<br>> | #define HAVE_GLOB 1
<br>> | #define HAVE_ATEXIT 1
<br>> | #define HAVE_BASENAME 1
<br>> | #define HAVE_BCOPY 1
<br>> | #define HAVE_BZERO 1
<br>> | #define HAVE_CHMOD 1
<br>> | #define HAVE_DUP2 1
<br>> | #define HAVE_ENDGRENT 1
<br>> | #define HAVE_ENDPWENT 1
<br>> | #define HAVE_EXECVP 1
<br>> | #define HAVE_EXPM1 1
<br>> | #define HAVE_EXPM1F 1
<br>> | #define HAVE_FCNTL 1
<br>> | #define HAVE_FORK 1
<br>> | #define HAVE_GETCWD 1
<br>> | #define HAVE_GETEGID 1
<br>> | #define HAVE_GETEUID 1
<br>> | #define HAVE_GETGID 1
<br>> | #define HAVE_GETGRENT 1
<br>> | #define HAVE_GETGRGID 1
<br>> | #define HAVE_GETGRNAM 1
<br>> | #define HAVE_GETPGRP 1
<br>> | #define HAVE_GETPID 1
<br>> | #define HAVE_GETPPID 1
<br>> | #define HAVE_GETPWENT 1
<br>> | #define HAVE_GETPWUID 1
<br>> | #define HAVE_GETTIMEOFDAY 1
<br>> | #define HAVE_GETUID 1
<br>> | #define HAVE_GETWD 1
<br>> | #define HAVE_KILL 1
<br>> | #define HAVE_LGAMMA 1
<br>> | #define HAVE_LGAMMAF 1
<br>> | #define HAVE_LGAMMA_R 1
<br>> | #define HAVE_LGAMMAF_R 1
<br>> | #define HAVE_LINK 1
<br>> | #define HAVE_LOCALTIME_R 1
<br>> | #define HAVE_LOG1P 1
<br>> | #define HAVE_LOG1PF 1
<br>> | #define HAVE_LSTAT 1
<br>> | #define HAVE_MEMMOVE 1
<br>> | #define HAVE_MKDIR 1
<br>> | #define HAVE_MKFIFO 1
<br>> | #define HAVE_MKSTEMP 1
<br>> | #define HAVE_ON_EXIT 1
<br>> | #define HAVE_PIPE 1
<br>> | #define HAVE_POLL 1
<br>> | #define HAVE_PUTENV 1
<br>> | #define HAVE_RAISE 1
<br>> | #define HAVE_READLINK 1
<br>> | #define HAVE_REALPATH 1
<br>> | #define HAVE_RENAME 1
<br>> | #define HAVE_RINDEX 1
<br>> | #define HAVE_RMDIR 1
<br>> | #define HAVE_ROUND 1
<br>> | #define HAVE_SELECT 1
<br>> | #define HAVE_SETGRENT 1
<br>> | #define HAVE_SETLOCALE 1
<br>> | #define HAVE_SETPWENT 1
<br>> | #define HAVE_SETVBUF 1
<br>> | #define HAVE_SIGACTION 1
<br>> | #define HAVE_SIGPENDING 1
<br>> | #define HAVE_SIGPROCMASK 1
<br>> | #define HAVE_SIGSUSPEND 1
<br>> | #define HAVE_SNPRINTF 1
<br>> | #define HAVE_STAT 1
<br>> | #define HAVE_STRCASECMP 1
<br>> | #define HAVE_STRDUP 1
<br>> | #define HAVE_STRERROR 1
<br>> | #define HAVE_STRNCASECMP 1
<br>> | #define HAVE_STRPTIME 1
<br>> | #define HAVE_STRSIGNAL 1
<br>> | #define HAVE_SYMLINK 1
<br>> | #define HAVE_TEMPNAM 1
<br>> | #define HAVE_TGAMMA 1
<br>> | #define HAVE_TGAMMAF 1
<br>> | #define HAVE_TRUNC 1
<br>> | #define HAVE_UMASK 1
<br>> | #define HAVE_UNAME 1
<br>> | #define HAVE_UNLINK 1
<br>> | #define HAVE_USLEEP 1
<br>> | #define HAVE_UTIME 1
<br>> | #define HAVE_VFPRINTF 1
<br>> | #define HAVE_VSPRINTF 1
<br>> | #define HAVE_VSNPRINTF 1
<br>> | #define HAVE_WAITPID 1
<br>> | #define HAVE__CHMOD 1
<br>> | #define HAVE__SNPRINTF 1
<br>> | #define HAVE_STRFTIME 1
<br>> | #define OCTAVE_HAVE_BROKEN_STRPTIME 1
<br>> | #define HAVE_LIBDL 1
<br>> | #define HAVE_DLOPEN 1
<br>> | #define HAVE_DLSYM 1
<br>> | #define HAVE_DLERROR 1
<br>> | #define HAVE_DLCLOSE 1
<br>> | #define HAVE_DLOPEN_API 1
<br>> | #define ENABLE_DYNAMIC_LINKING 1
<br>> | #define HAVE_TIMEVAL 1
<br>> | #define HAVE_FINITE 1
<br>> | #define HAVE_ISNAN 1
<br>> | #define HAVE_ISINF 1
<br>> | #define HAVE_COPYSIGN 1
<br>> | #define HAVE__FINITE 1
<br>> | #define HAVE__ISNAN 1
<br>> | #define HAVE__COPYSIGN 1
<br>> | #define HAVE_DECL_SIGNBIT 1
<br>> | #define HAVE_ACOSH 1
<br>> | #define HAVE_ACOSHF 1
<br>> | #define HAVE_ASINH 1
<br>> | #define HAVE_ASINHF 1
<br>> | #define HAVE_ATANH 1
<br>> | #define HAVE_ATANHF 1
<br>> | #define HAVE_ERF 1
<br>> | #define HAVE_ERFF 1
<br>> | #define HAVE_ERFC 1
<br>> | #define HAVE_ERFCF 1
<br>> | #define HAVE_EXP2 1
<br>> | #define HAVE_EXP2F 1
<br>> | /* end confdefs.h. */
<br>> | /* Define log2 to an innocuous variant, in case <limits.h> declares log2.
<br>> | For example, HP-UX 11i <limits.h> declares gettimeofday. */
<br>> | #define log2 innocuous_log2
<br>> |
<br>> | /* System header to define __stub macros and hopefully few prototypes,
<br>> | which can conflict with char log2 (); below.
<br>> | Prefer <limits.h> to <assert.h> if __STDC__ is defined, since
<br>> | <limits.h> exists even on freestanding compilers. */
<br>> |
<br>> | #ifdef __STDC__
<br>> | # include <limits.h>
<br>> | #else
<br>> | # include <assert.h>
<br>> | #endif
<br>> |
<br>> | #undef log2
<br>> |
<br>> | /* Override any GCC internal prototype to avoid an error.
<br>> | Use char because int might match the return type of a GCC
<br>> | builtin and then its argument prototype would still apply. */
<br>> | #ifdef __cplusplus
<br>> | extern "C"
<br>> | #endif
<br>> | char log2 ();
<br>> | /* The GNU C library defines this for functions which it implements
<br>> | to always fail with ENOSYS. Some functions are actually named
<br>> | something starting with __ and the normal name is an alias. */
<br>> | #if defined __stub_log2 || defined __stub___log2
<br>> | choke me
<br>> | #endif
<br>> |
<br>> | #ifdef F77_DUMMY_MAIN
<br>> |
<br>> | # ifdef __cplusplus
<br>> | extern "C"
<br>> | # endif
<br>> | int F77_DUMMY_MAIN() { return 1; }
<br>> |
<br>> | #endif
<br>> | int
<br>> | main ()
<br>> | {
<br>> | return log2 ();
<br>> | ;
<br>> | return 0;
<br>> | }
<br>>
<br>>
<br>> I think that the lines like the following will be required
<br>>
<br>> #ifndef HAVE_LOG2
<br>> #define log2(x) xlog2(x)
<br>> #endif
<br>>
</div><br>Almost certainly not. Actually some hidden macro like this interferes
<br>with the octave_value
<br>method definitions.
<br><br><div class='shrinkable-quote'><br>> Regards
<br>>
<br>> Tatsuro
<br>>
<br>>
<br>>
<br>>
<br>> --------------------------------------
<br>> Stop! Global Warming ~ Yahoo! JAPAN Earth Project
<br>> <a href="http://pr.mail.yahoo.co.jp/earthproject/" target="_top" rel="nofollow">http://pr.mail.yahoo.co.jp/earthproject/</a><br>>
</div><br><br><br>--
<br>RNDr. Jaroslav Hajek
<br>computing expert
<br>Aeronautical Research and Test Institute (VZLU)
<br>Prague, Czech Republic
<br>url: www.highegg.matfyz.cz
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18632145&i=2" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18631804Re: about axes properties2008-07-24T05:34:58Z2008-07-24T05:34:58ZJohn W. EatonOn 24-Jul-2008, Francesco Potorti` wrote:
<br><br>| >My understanding is that the parameter for outerposition is a vector
<br>| >p(1,2,3,4) with:
<br>| >
<br>| >p(1) is the x-coordinate of the lower left corner of the plot
<br>| >p(2) is the y-coordinate of the lower left corner of the plot
<br>| >p(3) is the width of the plot
<br>| >p(4) is the height of the plot
<br>| >
<br>| >All these values are given in relative units, i.e. in fractions of the
<br>| >width and height of the graph window, where (0/0) is the lower left
<br>| >corner of the window and (1/1) is the upper right corner.
<br>|
<br>| Yes, this is I too deduced from the current documentation.
<br>|
<br>| >Does that make sense?
<br>|
<br>| Yes, this is approximately the behaviour I observe.
<br>|
<br>| >If so, something along these lines should propably be added to the
<br>| >documentation.
<br>|
<br>| I could do that, but some things are not clear to me.
<br>|
<br>| 1) I can enlarge the plot up to 1.2 without exceeding the window
<br>| boundaries.
<br>|
<br>| 2) Apparently, when you make subplots, you can still move them around
<br>| through the whole window, which is good and useful, but also means
<br>| that "outerposition" is somehow initialised: why can't one look at
<br>| the initial values?
<br>|
<br>| 3) How does "aspectratio" interact with the imposed size?
<br><br><br>Please remember that with the current graphics system, we are
<br>generally doing what we can to approximate the relatively large and
<br>complex Matlab graphics system given limited developer resources and
<br>the limitations of gnuplot. It might not always be possible to do the
<br>Matlab-compatible thing given gnuplot's features. Also, we are in the
<br>process of replacing the gnuplot backend with an OpenGL-based
<br>renderer, so any heroic efforts to improve the gnuplot backend might
<br>not be justified. If you'd like to discuss this further, or help to
<br>improve Octave's graphics system, I think the maintainers list is a
<br>better place for the discussion..
<br><br>jwe
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18631804&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18631459Re: make error about log2 for octave-3.1.51 (Not 3.1.50)2008-07-24T05:30:09Z2008-07-24T05:30:09ZJohn W. EatonOn 24-Jul-2008, Tatsuro MATSUOKA wrote:
<br><br>| Hello Jaroslav
<br>|
<br>| Thank you for your reply
<br>|
<br>|
<br>| > This looks really weird given that there is no / token on the line. I
<br>| > suspect that
<br>| > you have log2 defined to some macro and that causes problems.
<br>|
<br>| I do not define log2 myself. Perhaps gcc on cygwin defines it.
<br><br>Can you please find out where and how it is defined? There is nothing
<br>in the Octave source code that #defines it.
<br><br>| I think that the lines like the following will be required
<br>|
<br>| #ifndef HAVE_LOG2
<br>| #define log2(x) xlog2(x)
<br>| #endif
<br><br>Where?
<br><br>I think it might be more appropriate to #undef log2 at some point(s),
<br>similar to what we do for min/max on Windows systems.
<br><br>In any case, if you think this is a bug in Octave that requires a fix,
<br>we should probably be discussing it on the bug list, not here.
<br><br>jwe
<br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18631459&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18630045Re: datestr bug [changeset]2008-07-24T04:28:03Z2008-07-24T04:28:03ZBen Abbott<br>On Jul 23, 2008, at 8:15 PM, Ben Abbott wrote:
<br><div class='shrinkable-quote'><div class='shrinkable-quote'><br>> On Wednesday, July 23, 2008, at 07:46PM, "Rob Mahurin" <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18630045&i=0" target="_top" rel="nofollow">rob@...</a>>
<br>> wrote:
<br>>> On Jul 23, 2008, at 5:47 PM, John W. Eaton wrote:
<br>>>> On 23-Jul-2008, Rob Mahurin wrote:
<br>>>>
<br>>>> | Hello, bug-octave.
<br>>>> |
<br>>>> | This statement
<br>>>> |
<br>>>> | > str = datestr(day, "yyyy-mm-dd\ndddd\n");
<br>>>> | > printf(str);
<br>>>> |
<br>>>> | seems to do some extra substituting:
<br>>>> |
<br>>>> | > 2009-03-14
<br>>>> | > SaturSay
<br>>>> |
<br>>>> | > 2008-11-26
<br>>>> | > WeWnesWay
<br>>>> |
<br>>>> |
<br>>>> | I'm running 3.0.1
<br>>>>
<br>>>> I think this has been reported at least once before. See for
<br>>>> example
<br>>>>
<br>>>> <a href="http://www.nabble.com/datestr-weekday-strptime%2C-2-bugs--" target="_top">http://www.nabble.com/datestr-weekday-strptime%2C-2-bugs--</a><br>>>> to7650360.html#a7650360
<br>>>>
<br>>>> I'm not sure what the best fix is. I would consider a patch.
<br>>>
<br>>> It seems like the offending line is
<br>>>
<br>>>> str = strrep (str, "d", names_d{weekday (datenum (v(i,1), v(i,2), v
<br>>>> (i,3)))});
<br>>>
<br>>>
<br>>> in datestr.m, which appears after the "dddd" in-place substitution.
<br>>> A possibly-good-enough patch would be to follow this with
<br>>>
<br>>>> # incomplete fix for greedy matching
<br>>>> str = strrep (str, "SunSay", "Sunday");
<br>>>> str = strrep (str, "MonMay", "Monday");
<br>>>> ...
<br>>>
<br>>>
<br>>> I think the "most correct" way to do this would be a set of regular
<br>>> expressions to identify the location and type of each format
<br>>> specifier, followed by a single-pass substitution.
<br>>>
<br>>> If there were a utility function with the right magic numbers in it I
<br>>> could use strftime, which has more robust formatting anyway. This:
<br>>>
<br>>>> printf(strftime ("%A %F %H:%M:%S %Z\n", localtime((now - 719529)
<br>>>> *24*3600)))
<br>>>
<br>>>
<br>>> (based on the hint from "help now") seems to go the wrong way from
<br>>> UTC.
<br>>>
<br>>> Cheers,
<br>>> Rob
<br>>
<br>> Another possible solution (hack) is to replace the offending line with
<br>>
<br>> nd = strfind (str, "d");
<br>> nd = setdiff (nd, strfind (str,"day"));
<br>> nd = setdiff (nd, strfind (str,"dnesday"));
<br>> str(nd) = names_d{weekday (datenum (v(i,1), v(i,2), v(i,3)))};
<br>>
<br>> In either event, an additional test should be added.
<br>>
<br>> %!assert(datestr(testtime+[0 0 3 0 0 0],"dddd"),"Wednesday")
<br>>
<br>> Comments?
<br>>
<br>> Ben
</div></div>Please consider the attached changeset.
<br><br><br /># HG changeset patch
<br># User Ben Abbott <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18630045&i=1" target="_top" rel="nofollow">bpabbott@...</a>>
<br># Date 1216898789 14400
<br># Node ID 738a8cbe661439b4eba7492a448bcd5b43faeb3a
<br># Parent 6add0f974aee3bf641842a71472a6014fdf4f442
<br>datestr.m: Fix format "dddd".
<br><br>diff -r 6add0f974aee -r 738a8cbe6614 scripts/ChangeLog
<br>--- a/scripts/ChangeLog Tue Jul 22 17:24:17 2008 -0400
<br>+++ b/scripts/ChangeLog Thu Jul 24 07:26:29 2008 -0400
<br>@@ -1,3 +1,7 @@
<br>+2008-07-23 Ben Abbott <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18630045&i=2" target="_top" rel="nofollow">bpabbott@...</a>>
<br>+
<br>+ * time/datestr.m: Fix format "dddd".
<br>+
<br> 2008-07-21 Michael Goffioul <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18630045&i=3" target="_top" rel="nofollow">michael.goffioul@...</a>>
<br>
<br> * plot/closereq.m: Use first gcbf to get the handle of the figure to
<br>diff -r 6add0f974aee -r 738a8cbe6614 scripts/time/datestr.m
<br>--- a/scripts/time/datestr.m Tue Jul 22 17:24:17 2008 -0400
<br>+++ b/scripts/time/datestr.m Thu Jul 24 07:26:29 2008 -0400
<br>@@ -275,7 +275,10 @@
<br> str = strrep (str, "dddd", names_dddd{weekday (datenum (v(i,1), v(i,2), v(i,3)))});
<br> str = strrep (str, "ddd", names_ddd{weekday (datenum (v(i,1), v(i,2), v(i,3)))});
<br> str = strrep (str, "dd", sprintf ("%02d", v(i,3)));
<br>- str = strrep (str, "d", names_d{weekday (datenum (v(i,1), v(i,2), v(i,3)))});
<br>+ nd = strfind (str, "d");
<br>+ nd = setdiff (nd, strfind (str,"day"));
<br>+ nd = setdiff (nd, strfind (str,"dnesday"));
<br>+ str(nd) = names_d{weekday (datenum (v(i,1), v(i,2), v(i,3)))};
<br> ## replace time symbols with actual values
<br> if ((any (strfind (str, "AM")) || any (strfind (str, "PM"))))
<br> if (v(i,4) > 12)
<br>@@ -341,6 +344,7 @@
<br> %!assert(datestr(testtime,29),"20051218");
<br> %!assert(datestr(testtime,30),"20051218T023317");
<br> %!assert(datestr(testtime,31),"2005-12-18 02:33:17");
<br>+%!assert(datestr(testtime+[0 0 3 0 0 0],"dddd"),"Wednesday")
<br> ## avoid the bug where someone happens to give a vector of datenums that
<br> ## happens to be 6 wide
<br> %!assert(datestr(733452.933:733457.933), ["14-Feb-2008 22:23:31";"15-Feb-2008 22:23:31";"16-Feb-2008 22:23:31";"17-Feb-2008 22:23:31";"18-Feb-2008 22:23:31";"19-Feb-2008 22:23:31"])
<br><br /><br><br><br><br><br />_______________________________________________
<br>Bug-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18630045&i=4" target="_top" rel="nofollow">Bug-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/bug-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/bug-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---Bugs-f1896.html" embed="fixTarget[1896]" target="_top" >Octave - Bugs</a></p>tag:www.nabble.com,2006:post-18628856Re: make error about log2 for octave-3.1.51 (Not 3.1.50)2008-07-24T02:54:02Z2008-07-24T02:54:02ZTatsuro MATSUOKA-2Hello Jaroslav
<br><br>Thank you for your reply
<br><br><br>> This looks really weird given that there is no / token on the line. I
<br>> suspect that
<br>> you have log2 defined to some macro and that causes problems.
<br><br>I do not define log2 myself. Perhaps gcc on cygwin defines it.
<br>Since octave 3.0.1, the log2 fuction has no problem at building.
<br><br><br><br>> Can you try running:
<br>>
<br>> /opt/octave/gcc-3.4.4dw2/bin/g++ -E -C
<br>> -I/opt/octave/gcc-3.4.4dw2/include -I/usr/include
<br>> -I/opt/octave/octave-3.1.51/include -I. -I.. -I../liboctave -I../src
<br>> -I../libcruft/misc
<br>> -DHAVE_CONFIG_H -mieee-fp -Wall -W -Wshadow -Wold-style-cast -g -O2 Cell.cc
<br>>
<br>> in the directory where the make failure occured and posting the output?
<br>I will show the results of which the log2 related.
<br><br>493: /* Define to 1 if you have the `log2' function. */
<br>494: /* #undef HAVE_LOG2 */
<br><br><br>The below lines perhaps are not related but I will show them as a reference
<br>47686: extern double xlog2 (double x);
<br>47687: extern Complex xlog2 (const Complex& x);
<br>47688: extern double xlog2 (double x, int& exp);
<br>47689: extern Complex xlog2 (const Complex& x, int& exp);
<br><br>47740: extern float xlog2 (float x);
<br>47741: extern FloatComplex xlog2 (const FloatComplex& x);
<br>47742: extern float xlog2 (float x, int& exp);
<br>47743: extern FloatComplex xlog2 (const FloatComplex& x, int& exp);
<br><br>To see the how configure detect log2
<br>configure:25965: checking for log2
<br>configure:26029: /opt/octave/gcc-3.4.4dw2/bin/gcc -o conftest.exe -g -O2
<br>-I/opt/octave/gcc-3.4.4dw2/include -I/usr/include -I/opt/octave/octave-3.1.51/include
<br>-L/opt/octave/gcc-3.4.4dw2/lib -L/opt/octave/octave-3.1.51/lib conftest.c -ldl -lwsock32 -lblas -lhdf5
<br>-lz -lm -lwsock32 >&5
<br>conftest.c:265: warning: conflicting types for built-in function 'log2'
<br>/cygdrive/c/DOCUME~1/Tatsu/LOCALS~1/Temp/cccEhDNQ.o: In function `main':
<br>/home/octaves/octave-3.1.51/conftest.c:284: undefined reference to `_log2'
<br>collect2: ld returned 1 exit status
<br>configure:26035: $? = 1
<br>configure: failed program was:
<br>| /* confdefs.h. */
<br>| #define PACKAGE_NAME ""
<br>| #define PACKAGE_TARNAME ""
<br>| #define PACKAGE_VERSION ""
<br>| #define PACKAGE_STRING ""
<br>| #define PACKAGE_BUGREPORT ""
<br>| #define OCTAVE_SOURCE 1
<br>| #define _GNU_SOURCE 1
<br>| #define STDC_HEADERS 1
<br>| #define HAVE_SYS_TYPES_H 1
<br>| #define HAVE_SYS_STAT_H 1
<br>| #define HAVE_STDLIB_H 1
<br>| #define HAVE_STRING_H 1
<br>| #define HAVE_MEMORY_H 1
<br>| #define HAVE_STRINGS_H 1
<br>| #define HAVE_INTTYPES_H 1
<br>| #define HAVE_STDINT_H 1
<br>| #define HAVE_UNISTD_H 1
<br>| #define SEPCHAR ':'
<br>| #define SEPCHAR_STR ":"
<br>| #define __NO_MATH_INLINES 1
<br>| #define CXX_NEW_FRIEND_TEMPLATE_DECL 1
<br>| #define CXX_ISO_COMPLIANT_LIBRARY 1
<br>| #define CXX_BROKEN_REINTERPRET_CAST 1
<br>| #define CXX_ABI gnu_v3
<br>| #define HAVE_LIBM 1
<br>| #define HAVE_QHULL 1
<br>| #define HAVE_PCRE 1
<br>| #define HAVE_REGEXEC 1
<br>| #define HAVE_REGEX 1
<br>| #define HAVE_ZLIB_H 1
<br>| #define HAVE_ZLIB 1
<br>| #define HAVE_HDF5_H 1
<br>| #define HAVE_HDF5 1
<br>| #define HAVE_H5GGET_NUM_OBJS 1
<br>| #define HAVE_FFTW3 1
<br>| #define HAVE_GLPK_H 1
<br>| #define HAVE_GLPK 1
<br>| #define HAVE_CURL_CURL_H 1
<br>| #define HAVE_CURL 1
<br>| #define HAVE_MAGICK 1
<br>| #define HAVE_GL_GL_H 1
<br>| #define HAVE_GL_GLU_H 1
<br>| #define HAVE_OPENGL 1
<br>| #define HAVE_FLTK 1
<br>| #define HAVE_IEEE754_DATA_FORMAT 1
<br>| #define F77_FUNC(name,NAME) name ## _
<br>| #define F77_FUNC_(name,NAME) name ## __
<br>| #define HAVE_BLAS 1
<br>| #define HAVE_SUITESPARSE_AMD_H 1
<br>| #define HAVE_AMD 1
<br>| #define HAVE_SUITESPARSE_UMFPACK_H 1
<br>| #define HAVE_UMFPACK 1
<br>| #define UMFPACK_SEPARATE_SPLIT 1
<br>| #define HAVE_SUITESPARSE_COLAMD_H 1
<br>| #define HAVE_COLAMD 1
<br>| #define HAVE_SUITESPARSE_CCOLAMD_H 1
<br>| #define HAVE_CCOLAMD 1
<br>| #define HAVE_SUITESPARSE_CHOLMOD_H 1
<br>| #define HAVE_CHOLMOD 1
<br>| #define HAVE_SUITESPARSE_CS_H 1
<br>| #define HAVE_CXSPARSE 1
<br>| #define HAVE_GETHOSTNAME 1
<br>| #define HAVE_GETPWNAM 1
<br>| #define HAVE_LIBWSOCK32 1
<br>| #define HAVE_DEV_T 1
<br>| #define HAVE_INO_T 1
<br>| #define HAVE_NLINK_T 1
<br>| #define HAVE_NLINK_T 1
<br>| #define HAVE_LONG_LONG_INT 1
<br>| #define HAVE_UNSIGNED_LONG_LONG_INT 1
<br>| #define HAVE_SIGSET_T 1
<br>| #define HAVE_SIG_ATOMIC_T 1
<br>| #define SIZEOF_SHORT 2
<br>| #define SIZEOF_INT 4
<br>| #define SIZEOF_LONG 4
<br>| #define SIZEOF_LONG_LONG 8
<br>| #define HAVE_ALLOCA_H 1
<br>| #define HAVE_ALLOCA 1
<br>| #define NPOS std::string::npos
<br>| #define HAVE_PLACEMENT_DELETE 1
<br>| #define HAVE_DYNAMIC_AUTO_ARRAYS 1
<br>| #define STDC_HEADERS 1
<br>| #define HAVE_DIRENT_H 1
<br>| #define TIME_WITH_SYS_TIME 1
<br>| #define HAVE_SYS_WAIT_H 1
<br>| #define HAVE_ASSERT_H 1
<br>| #define HAVE_CURSES_H 1
<br>| #define HAVE_DLFCN_H 1
<br>| #define HAVE_FCNTL_H 1
<br>| #define HAVE_FLOAT_H 1
<br>| #define HAVE_GRP_H 1
<br>| #define HAVE_IEEEFP_H 1
<br>| #define HAVE_INTTYPES_H 1
<br>| #define HAVE_LIMITS_H 1
<br>| #define HAVE_LOCALE_H 1
<br>| #define HAVE_MEMORY_H 1
<br>| #define HAVE_NCURSES_H 1
<br>| #define HAVE_POLL_H 1
<br>| #define HAVE_PTHREAD_H 1
<br>| #define HAVE_PWD_H 1
<br>| #define HAVE_STDINT_H 1
<br>| #define HAVE_STDLIB_H 1
<br>| #define HAVE_STRING_H 1
<br>| #define HAVE_SYS_IOCTL_H 1
<br>| #define HAVE_SYS_PARAM_H 1
<br>| #define HAVE_SYS_POLL_H 1
<br>| #define HAVE_SYS_RESOURCE_H 1
<br>| #define HAVE_SYS_SELECT_H 1
<br>| #define HAVE_SYS_STAT_H 1
<br>| #define HAVE_SYS_TIME_H 1
<br>| #define HAVE_SYS_TIMES_H 1
<br>| #define HAVE_SYS_TYPES_H 1
<br>| #define HAVE_SYS_UTSNAME_H 1
<br>| #define HAVE_SYS_UTIME_H 1
<br>| #define HAVE_TERMCAP_H 1
<br>| #define HAVE_UNISTD_H 1
<br>| #define HAVE_UTIME_H 1
<br>| #define HAVE_SSTREAM 1
<br>| #define HAVE_TERMIO_H 1
<br>| #define HAVE_GLOB_H 1
<br>| #define HAVE_FNMATCH_H 1
<br>| #define HAVE_FNMATCH 1
<br>| #define HAVE_GLOB 1
<br>| #define HAVE_ATEXIT 1
<br>| #define HAVE_BASENAME 1
<br>| #define HAVE_BCOPY 1
<br>| #define HAVE_BZERO 1
<br>| #define HAVE_CHMOD 1
<br>| #define HAVE_DUP2 1
<br>| #define HAVE_ENDGRENT 1
<br>| #define HAVE_ENDPWENT 1
<br>| #define HAVE_EXECVP 1
<br>| #define HAVE_EXPM1 1
<br>| #define HAVE_EXPM1F 1
<br>| #define HAVE_FCNTL 1
<br>| #define HAVE_FORK 1
<br>| #define HAVE_GETCWD 1
<br>| #define HAVE_GETEGID 1
<br>| #define HAVE_GETEUID 1
<br>| #define HAVE_GETGID 1
<br>| #define HAVE_GETGRENT 1
<br>| #define HAVE_GETGRGID 1
<br>| #define HAVE_GETGRNAM 1
<br>| #define HAVE_GETPGRP 1
<br>| #define HAVE_GETPID 1
<br>| #define HAVE_GETPPID 1
<br>| #define HAVE_GETPWENT 1
<br>| #define HAVE_GETPWUID 1
<br>| #define HAVE_GETTIMEOFDAY 1
<br>| #define HAVE_GETUID 1
<br>| #define HAVE_GETWD 1
<br>| #define HAVE_KILL 1
<br>| #define HAVE_LGAMMA 1
<br>| #define HAVE_LGAMMAF 1
<br>| #define HAVE_LGAMMA_R 1
<br>| #define HAVE_LGAMMAF_R 1
<br>| #define HAVE_LINK 1
<br>| #define HAVE_LOCALTIME_R 1
<br>| #define HAVE_LOG1P 1
<br>| #define HAVE_LOG1PF 1
<br>| #define HAVE_LSTAT 1
<br>| #define HAVE_MEMMOVE 1
<br>| #define HAVE_MKDIR 1
<br>| #define HAVE_MKFIFO 1
<br>| #define HAVE_MKSTEMP 1
<br>| #define HAVE_ON_EXIT 1
<br>| #define HAVE_PIPE 1
<br>| #define HAVE_POLL 1
<br>| #define HAVE_PUTENV 1
<br>| #define HAVE_RAISE 1
<br>| #define HAVE_READLINK 1
<br>| #define HAVE_REALPATH 1
<br>| #define HAVE_RENAME 1
<br>| #define HAVE_RINDEX 1
<br>| #define HAVE_RMDIR 1
<br>| #define HAVE_ROUND 1
<br>| #define HAVE_SELECT 1
<br>| #define HAVE_SETGRENT 1
<br>| #define HAVE_SETLOCALE 1
<br>| #define HAVE_SETPWENT 1
<br>| #define HAVE_SETVBUF 1
<br>| #define HAVE_SIGACTION 1
<br>| #define HAVE_SIGPENDING 1
<br>| #define HAVE_SIGPROCMASK 1
<br>| #define HAVE_SIGSUSPEND 1
<br>| #define HAVE_SNPRINTF 1
<br>| #define HAVE_STAT 1
<br>| #define HAVE_STRCASECMP 1
<br>| #define HAVE_STRDUP 1
<br>| #define HAVE_STRERROR 1
<br>| #define HAVE_STRNCASECMP 1
<br>| #define HAVE_STRPTIME 1
<br>| #define HAVE_STRSIGNAL 1
<br>| #define HAVE_SYMLINK 1
<br>| #define HAVE_TEMPNAM 1
<br>| #define HAVE_TGAMMA 1
<br>| #define HAVE_TGAMMAF 1
<br>| #define HAVE_TRUNC 1
<br>| #define HAVE_UMASK 1
<br>| #define HAVE_UNAME 1
<br>| #define HAVE_UNLINK 1
<br>| #define HAVE_USLEEP 1
<br>| #define HAVE_UTIME 1
<br>| #define HAVE_VFPRINTF 1
<br>| #define HAVE_VSPRINTF 1
<br>| #define HAVE_VSNPRINTF 1
<br>| #define HAVE_WAITPID 1
<br>| #define HAVE__CHMOD 1
<br>| #define HAVE__SNPRINTF 1
<br>| #define HAVE_STRFTIME 1
<br>| #define OCTAVE_HAVE_BROKEN_STRPTIME 1
<br>| #define HAVE_LIBDL 1
<br>| #define HAVE_DLOPEN 1
<br>| #define HAVE_DLSYM 1
<br>| #define HAVE_DLERROR 1
<br>| #define HAVE_DLCLOSE 1
<br>| #define HAVE_DLOPEN_API 1
<br>| #define ENABLE_DYNAMIC_LINKING 1
<br>| #define HAVE_TIMEVAL 1
<br>| #define HAVE_FINITE 1
<br>| #define HAVE_ISNAN 1
<br>| #define HAVE_ISINF 1
<br>| #define HAVE_COPYSIGN 1
<br>| #define HAVE__FINITE 1
<br>| #define HAVE__ISNAN 1
<br>| #define HAVE__COPYSIGN 1
<br>| #define HAVE_DECL_SIGNBIT 1
<br>| #define HAVE_ACOSH 1
<br>| #define HAVE_ACOSHF 1
<br>| #define HAVE_ASINH 1
<br>| #define HAVE_ASINHF 1
<br>| #define HAVE_ATANH 1
<br>| #define HAVE_ATANHF 1
<br>| #define HAVE_ERF 1
<br>| #define HAVE_ERFF 1
<br>| #define HAVE_ERFC 1
<br>| #define HAVE_ERFCF 1
<br>| #define HAVE_EXP2 1
<br>| #define HAVE_EXP2F 1
<br>| /* end confdefs.h. */
<br>| /* Define log2 to an innocuous variant, in case <limits.h> declares log2.
<br>| For example, HP-UX 11i <limits.h> declares gettimeofday. */
<br>| #define log2 innocuous_log2
<br>|
<br>| /* System header to define __stub macros and hopefully few prototypes,
<br>| which can conflict with char log2 (); below.
<br>| Prefer <limits.h> to <assert.h> if __STDC__ is defined, since
<br>| <limits.h> exists even on freestanding compilers. */
<br>|
<br>| #ifdef __STDC__
<br>| # include <limits.h>
<br>| #else
<br>| # include <assert.h>
<br>| #endif
<br>|
<br>| #undef log2
<br>|
<br>| /* Override any GCC internal prototype to avoid an error.
<br>| Use char because int might match the return type of a GCC
<br>| builtin and then its argument prototype would still apply. */
<br>| #ifdef __cplusplus
<br>| extern "C"
<br>| #endif
<br>| char log2 ();
<br>| /* The GNU C library defines this for functions which it implements
<br>| to always fail with ENOSYS. Some functions are actually named
<br>| something starting with __ and the normal name is an alias. */
<br>| #if defined __stub_log2 || defined __stub___log2
<br>| choke me
<br>| #endif
<br>|
<br>| #ifdef F77_DUMMY_MAIN
<br>|
<br>| # ifdef __cplusplus
<br>| extern "C"
<br>| # endif
<br>| int F77_DUMMY_MAIN() { return 1; }
<br>|
<br>| #endif
<br>| int
<br>| main ()
<br>| {
<br>| return log2 ();
<br>| ;
<br>| return 0;
<br>| }
<br><br><br>I think that the lines like the following will be required
<br><br>#ifndef HAVE_LOG2
<br>#define log2(x) xlog2(x)
<br>#endif
<br><br>Regards
<br><br>Tatsuro
<br><br><br><br><br>--------------------------------------
<br>Stop! Global Warming ~ Yahoo! JAPAN Earth Project
<br><a href="http://pr.mail.yahoo.co.jp/earthproject/" target="_top" rel="nofollow">http://pr.mail.yahoo.co.jp/earthproject/</a><br>_______________________________________________
<br>Help-octave mailing list
<br><a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18628856&i=0" target="_top" rel="nofollow">Help-octave@...</a>
<br><a href="https://www.cae.wisc.edu/mailman/listinfo/help-octave" target="_top" rel="nofollow">https://www.cae.wisc.edu/mailman/listinfo/help-octave</a><br><p>From forum: <a href="http://www.nabble.com/Octave---General-f1897.html" embed="fixTarget[1897]" target="_top" >Octave - General</a></p>tag:www.nabble.com,2006:post-18628505Re: Octave 3.1.51 available for ftp2008-07-24T02:38:23Z2008-07-24T02:38:23ZRafael Laboissiere* John W. Eaton <<a href="http://www.nabble.com/user/SendEmail.jtp?type=post&post=18628505&i=0" target="_top" rel="nofollow">jwe@...</a>> [2008-07-22 11:39]:
<br><br>> A new snapshot of Octave is now available for ftp from ftp.octave.org in the
<br>> directory /pub/octave/bleeding-edge:
<br>> [snip]
<br><br>Debian packages for 3.1.51 have been uploaded to experimental [1] and are
<br>also available from my apt-getable repository [2], as usual.
<br>
<br>[1] <a href="http://ftp-master.debian.org/new/octave3.1_3.1.51-1.html" target="_top" rel="nofollow">http://ftp-master.debian.org/new/octave3.1_3.1.51-1.html</a><br>[2] <a href="http://people.debian.org/~rafael/octave3.1/" target="_top" rel="nofollow">http://people.debian.org/~rafael/octave3.1/</a><br><br>--
<br>Rafael
<br><p>From forum: <a href="http://www.nabble.com/Octave---Maintainers-f1900.html" embed="fixTarget[1900]" target="_top" >Octave - Maintainers</a></p>tag:www.nabble.com,2006:post-18627946Re: about axes properties2008-07-24T01:43:19Z2008-07-24T01:43:19ZFrancesco Potorti`-2<div class='shrinkable-quote'>>My understanding is that the parameter for outerposition is a vector
<br>>p(1,2,3,4) with:
<br>>
<br>>p(1) is the x-coordinate of the lower left corner of the plot
<br>>p(2) is the y-coordinate of the lower left corner of the plot
<br>>p(3) is the width of the plot
<br>>p(4) is the height of the plot
<br>>
<br>>All these values are given in relative units, i.e. in fractions of the
<br>>width and height of the graph window, where (0/0) is the lower left
<br>>corner of the window and (1/1) is the upper right corner.
</div><br>Yes, this is I too deduced from the current documentation.
<br><br>>Does that make sense?
<br&