Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.



Re: Delaunay triangulation, Maybe a bug
Posted:
Jun 10, 2014 2:04 AM


"h" wrote in message <ln5rso$ofb$1@newscl01ah.mathworks.com>... > Hello everyone, > > I use DT = delaunayTriangulation(node) to generate triangles. When I use 15*15 uniform distributed nodes, i get 404 triangular elements. If i am doing the right thing, i think the right answer should be 392 elements. > I calculated the area of these triangles and I found the area of some triangles are zero. > Maybe this is a bug. Hope someone can fix it.
I won't consider as BUG, after all delaunay triangulation is a _numerical_ calculation and a regular grid is a singular case for delaunay triangulation: if one of the node move slightly from it's ideal position, the area of the zerosurface becomes > 0.
And I'm not even sure your grid points coordinates can even stored with 100 % accuracy with floating points at the first place.
You just have to deal with this kind of fact in numerical calculation.
Bruno



