Skip to content
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

[REVIEW] Replace UMAP functionality dependency on nvgraph with RAFT Spectral Clustering #2500

Merged
merged 12 commits into from
Jul 17, 2020

Conversation

aschaffer
Copy link
Contributor

This is in conjunction with Spectral Clustering functionality move to RAFT (rapidsai/raft#12) and cugraph corresponding refactoring (rapidsai/cugraph#980).

@aschaffer aschaffer requested review from a team as code owners June 30, 2020 21:43
@GPUtester
Copy link
Contributor

Please update the changelog in order to start CI tests.

View the gpuCI docs here.

Copy link
Member

@cjnolet cjnolet left a comment

Choose a reason for hiding this comment

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

Looks good in general. Now that we've modularized the spectral implementation, it would be nice to take this opportunity to make the cluster solver execution optional when all we need are the eigenvectors.

cpp/src_prims/sparse/spectral.cuh Outdated Show resolved Hide resolved
@cjnolet cjnolet added 4 - Waiting on Author Waiting for author to respond to review CUDA / C++ CUDA issue Tech Debt Issues related to debt labels Jul 2, 2020
@cjnolet cjnolet self-assigned this Jul 8, 2020
Copy link
Member

@cjnolet cjnolet left a comment

Choose a reason for hiding this comment

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

LGTM pending update of commit hash once corresponding spectral PR is merged on the RAFT side.

Would also be nice to see the no-op clustering strategy make its way into raft eventually. I'm okay keeping it here until someone else needs it though.

Copy link
Member

@cjnolet cjnolet left a comment

Choose a reason for hiding this comment

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

I think we can remove the NVgraph dependency from CMakeLists.txt, otherwise this looks great.

Copy link
Member

@teju85 teju85 left a comment

Choose a reason for hiding this comment

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

Thanks @aschaffer for this PR! Just a couple of minor queries.

cpp/cmake/Dependencies.cmake Outdated Show resolved Hide resolved
cpp/cmake/Dependencies.cmake Show resolved Hide resolved
using index_type = int;
using value_type = T;

raft::handle_t r_handle;
Copy link
Member

Choose a reason for hiding this comment

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

@divyegala JFYI. Assuming this PR goes first before yours, we'll have to update this logic to use the raft handle that comes inside our cumlHandle.

@aschaffer
Copy link
Contributor Author

rerun tests.

@dantegd
Copy link
Member

dantegd commented Jul 16, 2020

rerun tests

Copy link
Member

@cjnolet cjnolet left a comment

Choose a reason for hiding this comment

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

LGTM

@aschaffer aschaffer requested a review from teju85 July 16, 2020 21:39
@dantegd
Copy link
Member

dantegd commented Jul 16, 2020

@teju85 could you re-review/approve this PR? Its just missing your approval to be able to merge

@dantegd dantegd added 4 - Waiting on Reviewer Waiting for reviewer to review or respond and removed 4 - Waiting on Author Waiting for author to respond to review labels Jul 16, 2020
Copy link
Member

@teju85 teju85 left a comment

Choose a reason for hiding this comment

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

LGTM.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
4 - Waiting on Reviewer Waiting for reviewer to review or respond CUDA / C++ CUDA issue Tech Debt Issues related to debt
Projects
None yet
Development

Successfully merging this pull request may close these issues.

6 participants