Skip to content

WIP: Using DGL instead of Torch Geometric #48

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 33 commits into
base: main
Choose a base branch
from

Conversation

gennarinoos
Copy link
Collaborator

@gennarinoos gennarinoos commented Feb 4, 2025

  • Update and fix unit tests
  • Fix CI environment setup
  • Check if point sampler probabilities are negative
  • Double check arch changes (graphconv instead of devconv)
  • Re-enable surface distance loss
  • Re-enable surface distance loss
  • Double check algorithm complies with the one on the paper

@@ -38,7 +39,7 @@ def test_find_nearest_triangles(loss_fn):

nearest = loss_fn.find_nearest_triangles(vertices, faces)
assert nearest.shape[0] == faces.shape[0]
assert nearest.shape[1] == 1 # k-1 = 1, since k=2
assert nearest.shape[1] == k_val
Copy link
Owner

@martinnormark martinnormark Feb 7, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I believe the comment from before does not fit anymore. The function will only return less than k if the number of centroids are less than k. Before we found the k nearest and remove self connections - DGL's function always returns k number of neighbours regardless of exclude self.

Co-authored-by: Martin Høst Normark <m@martinnormark.com>
@martinnormark martinnormark mentioned this pull request Feb 12, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants