Excessive Vertex Duplication in cut_edges Function

[karinsifri]

When using the cut_edges function on a closed manifold mesh with a chain of connected edges, I noticed excessive vertex duplication. All edges in the mesh were duplicated, with most of the new vertices left unreferenced.

I followed the documentation-examples and used the following code:

import gpytoolbox as gpy

new_faces, new_vertex_idx = gpy.cut_edges(f, path_edges)
gpy.write_mesh("../meshes/debug.obj", v[new_vertex_idx, :], new_faces)

For example, the original mesh had 5389 vertices, but the length of the returned indices was 32322, even for a short path (of only 3 edges).

Upon inspecting the source code, I noticed that remove_unreferenced is already included. However, it does not appear to work effectively in reducing the excess vertices. I had to manually use:

new_v, new_f = gpy.remove_unreferenced(v[new_vertex_idx, :], new_faces)

While this workaround helps, the excessive duplication and the apparent ineffectiveness of remove_unreferenced make the process more cumbersome than necessary.

Let me know if more details or test cases are needed to help resolve the issue. I’d be happy to assist further.

[odedstein]

Hi @karinsifri ,
Can you suggest a fix (via PR) that includes the correct removing of unreferenced vertices?
(Please also include the example mesh you ran the code on.)

[karinsifri]

Thank you for your response! I appreciate the opportunity to contribute.

I’d be happy to look into the issue and attempt to debug the root cause in the code. While I currently need some time to review and understand the source of the problem, I’ll work on identifying the issue and suggesting a fix via a PR when I’m able.

In the meantime, I’ve included the details of the mesh I worked on and the path I was trying to cut, as shared earlier. I’ll keep you updated on my progress.

The mesh I have worked on is C_r2.txt (this is .obj file but i have changed it's extension to be able to upload it)

The path I was trying to cut looks something like

path_edges = np.array([[3460, 2852],
       [2852, 1939],
       [1939,  656],
       [ 656, 4535],
       [4535, 1866],
       [1866, 2327],
       [2327, 1814],
       [1814, 3020],
       [3020, 5060],
       [5060, 1661],
       [1661, 4912]], dtype=int64)