Tighter (albeit arbitrary) constraint on path end detection

This is really not great
This commit is contained in:
Grant Sanderson 2024-08-26 09:14:42 -05:00
parent 827f4db5e2
commit cd5c436ce4

View file

@ -726,7 +726,7 @@ class VMobject(Mobject):
return self.get_bezier_tuples_from_points(self.get_points())
def get_subpath_end_indices_from_points(self, points: Vect3Array) -> np.ndarray:
atol = self.tolerance_for_point_equality
atol = 1e-4 # TODO, this is too unsystematic
a0, h, a1 = points[0:-1:2], points[1::2], points[2::2]
# An anchor point is considered the end of a path
# if its following handle is sitting on top of it.