[Octave-bug-tracker] [bug #63704] isosurface returns incorrect results for data bigger than small

[Markus Mützel]

Update of bug #63704 (project octave):

                  Status:                    None => Confirmed              

    _______________________________________________________

Follow-up Comment #1:

I can confirm with Octave 8.0.90.
It also renders incorrectly when I change to `graphics_toolkit('fltk')` before
the test. So not much more information from that test.

Looking at the data of the patch graphics object:

>> hkids = get(gca, 'Children');
>> get(hkids, 'Type')
ans =
{
  [1,1] = light
  [2,1] = patch
}

>> hpatch = hkids(2);
>> vert = get(hpatch, 'Vertices');
>> magn = sqrt(sum(vert.^2,2));
>> find(max(abs(magn-1)) == abs(magn-1), 1, 'first')
ans = 29048
>> sqrt(sum(vert(29048,:).^2))
ans = 2.3583
>>


That's pretty far off from `1`...

There might be something wrong with the algorithm that calculates the vertices
for the isosurface (only if there are many edges?). But there haven't been
many changes to the involved functions between Octave 6 and Octave 7
afaict...



    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?63704>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/