From bug-octave-request at bevo dot che dot wisc dot edu Mon Apr  7 16:04:26 2003
Subject: It really is a bug in 2.1.46 (Was: Bug#187427: Bug in __bodquist__.m)
From: "John W. Eaton" <jwe at bevo dot che dot wisc dot edu>
To: Dirk Eddelbuettel <edd at debian dot org>
Cc: Mirek Kwasniak <mirek at zind dot ikem dot pwr dot wroc dot pl>, 187427@bugs.debian.org,    bug-octave@bevo.che.wisc.edu
Date: Mon, 7 Apr 2003 16:03:07 -0500

On  5-Apr-2003, Dirk Eddelbuettel <edd at debian dot org> wrote:

| On Sat, Apr 05, 2003 at 07:30:47PM +0200, Mirek Kwasniak wrote:
| > On Sat, Apr 05, 2003 at 09:50:42AM -0600, Dirk Eddelbuettel wrote:
| > [...]
| > > On Thu, Apr 03, 2003 at 09:35:39AM -0000, sacrificial-spam-address at horizon dot com wrote:
| > > > Package: octave2.1
| > > > Version: 2.1.46-3
| > > > 
| > > > John W. Easton wrote:
| > > > >| octave:1> bode(zp2sys([],[1,2],1))
| > > > >| error: logspace: arguments must be scalars
| > > > >| error: evaluating if command near line 71, column 5
| > > > >| error: evaluating if command near line 68, column 3
| > > > >| error: called from `logspace' in file
| > > > >| `/usr/share/octave/2.1.35/m/general/logspace.m'
| > > > >| error: called from `__bodquist__' in file
| > > > >| `/usr/share/octave/2.1.35/m/control/base/__bodquist__.m'
| > > > >| error: called from `bode' in file
| > > > >| `/usr/share/octave/2.1.35/m/control/base/bode.m'
| > > > >
| > > > > This problem seems to be fixed in 2.1.39.
| > > > 
| > > > It appears to be broken again in 2.1.46 (Debian binary package 2.1.46-3):
| > > > 
| > > > GNU Octave, version 2.1.46 (i386-pc-linux-gnu).
| > [...]
| > 
| > > Actually, could you please double check that your installation is fine?
| > > It works just dandily on my system:
| > > 
| > > edd at chibud:~> octave2.1
| > > GNU Octave, version 2.1.43 (i386-pc-linux-gnu).
| >                       ^^^^^^
| > 
| > Dirk, do you have 2.1.43? Bug is against 2.1.46-3 deb :)
| 
| Oh boy, I was obviously lacking coffee ...  I checked this on my testing
| machine knowing full well that I built more recently on unstable. Hence the
| version skew.
| 
| And you are of course correct in that this fails.
| 
| And it would appear that the culprit is not in the .m files as the /entire/
| control section is unchanged:
| 
| edd at chibud:~> diff -ru /usr/share/octave/2.1.43/m/control/ \
|     /var/local/unstable/usr/share/octave/2.1.46/m/control/
| edd at chibud:~>
| 
| So it must be due to some changes in the parse.  At which point I usually
| turn to John -- any idea at your end?
| 
| Dirk
| 
| > [...]
| > 
| > > Or could this may be a side-effect of me having octave-forge installed?
| > > 
| > > Please follow-up with additional information, or I'll have to close the bug
| > > report for apparent lack of reproducibility.
| > 
| > It's not octave-forge forge problem (I have it).
| > I think it's a problem in lines 105-112 in __bodquist__.m:
| > 105:       pdbig = vec(find(pd > dphase));
| >                    ^^^^
| > 110:       fdbig = vec(find(fd > fm/10));
| >                    ^^^
| > 112:      bigpts = union(fdbig, pdbig);
| > 
| > What is purpouse doing column vector? union works fine for rows.
| > Then we have a runtime error because in line:
| > 
| > 127:         for ii=bigpts
| > 
| > we get ii as column :(


I made the following changes.

Thanks,

jwe


scripts/ChangeLog:

2003-04-07  John W. Eaton  <jwe at bevo dot che dot wisc dot edu>

	* control/base/__bodquist__.m: Don't convert pdbig and fdbig to
	column vectors.


Index: control/base/__bodquist__.m
===================================================================
RCS file: /usr/local/cvsroot/octave/scripts/control/base/__bodquist__.m,v
retrieving revision 1.6
diff -u -r1.6 __bodquist__.m
--- control/base/__bodquist__.m	9 Aug 2002 18:58:13 -0000	1.6
+++ control/base/__bodquist__.m	7 Apr 2003 21:01:48 -0000
 at @ -102,12 +102,12 @@
     dmag = 0.2;         # desired max change in magnitude
     while(pcnt)
       pd = abs(diff(phase));                    # phase variation
-      pdbig = vec(find(pd > dphase));
+      pdbig = find(pd > dphase);
 
       lp = length(f);  lp1 = lp-1;              # relative variation
       fd = abs(diff(f));
       fm = max(abs([f(1:lp1); f(2:lp)]));
-      fdbig = vec(find(fd > fm/10));
+      fdbig = find(fd > fm/10);
 
       bigpts = union(fdbig, pdbig);
 



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------