[Gmsh] integrate bugs

[Lars Rindorf]

I solved the first 'bug'. I accidently integrated also over the boundaries as well as the open domain. The second bug still holds.
 
KR Lars

________________________________

From: Lars Rindorf
Sent: Tue 1/22/2008 10:02 AM
To: gmsh at geuz.org
Subject: integrate bugs


Dear Christophe
 
I found a couple of errors (I think) in the integrate plugin to gmesh. My OS is XP.
 
If I integrate '1' over a 1x1 square th plugin returns '3', if the square is 2x2 it returns '8'. It seems to follow a formula L^2 + 2*L, where L is the length of the side of the square. I also tried to integrate x^2 over 1x1 and it returned 1/4 instead of 1/12, wring by the factor of '3'.
 
There is a second error in the plugin 'Integrate'. Plot two scalars, first a and then b, integrate b and gmesh integrates a (the first), no matter if a or b is chosen in the view of the plugin.

Kind regards

Lars Rindorf