Open aramadancs opened 2 years ago
Thx for the bug report ! Let me ping @troussil on this one.
This is a patch of a fix I was testing locally. I assume it's related to the tolerance of Pi comparison. Hull2DHelpers.il.patch.txt
Let me ping @kerautret in my turn! I think he wrote the part about thickness computation.
(Note however that MelkmanConvexHull is guaranteed to properly work only with integer points, but not with real points due to possible inexact float comparisons.)
👍 thx @troussil, I thought it was from you.
Looking at the patch, the angle computation is a bit weird for just a sign of a determinant (even if computed on real numbers). @kerautret ?
Thanks for their report @aramadancs and the patch ! I look it asap
any update @kerautret ? thx
Done https://github.com/DGtal-team/DGtal/pull/1647 with some updates and including the example in tests.
The function returns 0 for a convex hull calculated using MelkmanConvexHull. I included the points of the convex hull directly.