From bug-octave-request at che dot utexas dot edu  Tue Aug 30 09:25:11 1994
From: Carlos Pfeiffer <pfeiffer at che dot utexas dot edu>
To: bug-octave
Date: Tue, 30 Aug 1994 09:25:09 -0500




I think I  found a bug on the function mesh.

I generate a "dome" with [xx,yy]=meshdom(x,y) and evaluated a function
on the points(xx(m,k),yy(m,k))


The matrixes xx and yy have dimensions "length(y) X length(x)".
Then the rows represent the "y" coordinate and the columns the
"x" coordinate.

However, when I use mesh(x,y,z) where z has dimensions equal to xx and yy,
it requires that the first parameter has the same number of elements as
the number of rows of z. Therefore I think it should be mesh(y,x,z).

Another problem is that the "y" coordinate is inverted according 
with the scale.

I fixed it by inverting the order of the rows of "z" premultiplying "z" by
a permutation matrix.

The example "sombrero" is symmetrical and has the same length of "x" and "y"
coordinates, so this "errors" are not noticed.

It also may be that I am misusing the mesh command, in this case I would
appreciate a comment on the right use.


Carlos F. Pfeiffer