| Thread Previous • Date Previous • Date Next • Thread Next |
1.3sec in average for triangulate() with the script lines commited earlier :-) (the other test was out-of-yade program). This is on intel Xeon. Let me know what it gives on your laptop if you give it a try. Note that it doesn't really include the cost of "tesselating" (computing dual points for all tetrahedra). Tesselation is computed silently inside computeVolumes(). I didn't measure exactly, but computeVolumes() takes less than one second in total and tesselation is really a small part of it. The major limitation of all this is you can't modelize complex or periodic boundaries (though multiple "infinite" planes could give a picewise planar approximate of any convex contour, and periodicity is available for the non-weighted Delaunay you've been using).Not yet. But "CGAL needs 1.4sec for 100k spheres" is very good. This is what CPU?
Bruno
| Thread Previous • Date Previous • Date Next • Thread Next |
