From owner-help-octave at bevo dot che dot wisc dot edu  Thu Nov  2 22:51:33 1995
Subject: mesh and meshdom bugs?
From: Joao Cardoso <jcardoso at bart dot inescn dot pt>
To: help-octave at bevo dot che dot wisc dot edu
Date: Fri, 3 Nov 1995 05:51:17 +0100 (MET)

Hi,

If I am correct, there is a bug in `meshdom' and `mesh'.

In `meshdom' for some strange reason the y vector is calculated backward:

     y = y (ylen:-1:1);         # jc: this line seems to be incorrect
     y = y (1:ylen);		# jc: it should(?) be this way

in `mesh', the following comparision also seems to be wrong:

      if (xlen == rows (z) && ylen == columns (z))	# wrong ?
      if (ylen == rows (z) && xlen == columns (z))	# right ?

The following "code" tryes to show both problems. It is the `sombrero' ploted
in the first quadrant. The first error message shows the error in `mesh'.
Now, watching the plot, one can see that the maximum of the function in not
at (0,0), showing the incorrection in `meshdom'.

octave:1> n=5;
x=linspace (0, 15, 5*n)';y=linspace (0,5, n)';
[xx, yy] = meshdom(x, y);
r = sqrt (xx .^ 2 + yy .^ 2) + eps;z = sin (r) ./ r;
mesh(x,y,z)

error: mesh: rows (z) must be the same as length (x) and
columns (z) must be the same as length (y)

octave:2> n=5;
x=linspace (0, 15, 5*n)';y=linspace (0,5, n)';
[xx, yy] = meshdom(x, y);
r = sqrt (xx .^ 2 + yy .^ 2) + eps;z = sin (r) ./ r;
mesh(x,y,z')		<--- note the transposition of z, for this to work

As I don't use this kind of ploting I am not absolutly shure of my reasons...

Thanks,

Joao