When you build a Delaunay_triangulation_3 in CGAL with points that are coplanar, you still get a nice 2D triangulation. However, when you build the corresponding Alpha_shape_3, CGAL (more precisely the function filtration()) claims that it is empty. The expected result is the filtration of the corresponding 2d alpha-complex (by the way, request for enhancement, implement Alpha_shape_2::filtration_with_alpha_values, preferably with a nicer interface than the 3D version).
Indeed, the various constructors have if (dimension() == 3) initialize_alpha(); and without initialize_alpha(); all the internal maps are empty and filtration returns immediately. Note that we cannot just run the initialize_alpha(); when dimension is 2, that segfaults.
In test/Alpha_shapes_3/include/CGAL/_test_weighted_alpha_shape_3.h, instead of the 13 points we currently insert, pick just
When you build a
Delaunay_triangulation_3
in CGAL with points that are coplanar, you still get a nice 2D triangulation. However, when you build the correspondingAlpha_shape_3
, CGAL (more precisely the functionfiltration()
) claims that it is empty. The expected result is the filtration of the corresponding 2d alpha-complex (by the way, request for enhancement, implement Alpha_shape_2::filtration_with_alpha_values, preferably with a nicer interface than the 3D version).Indeed, the various constructors have
if (dimension() == 3) initialize_alpha();
and withoutinitialize_alpha();
all the internal maps are empty andfiltration
returns immediately. Note that we cannot just run theinitialize_alpha();
when dimension is 2, that segfaults.In
test/Alpha_shapes_3/include/CGAL/_test_weighted_alpha_shape_3.h
, instead of the 13 points we currently insert, pick justand the test segfaults (I didn't check if that's the same issue).