Add support in C API for handling unweighted graphs in algorithms that expect weights #3513
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…s to support unweighted graphs
ChuckHastings
added
improvement
seunghwak
reviewed
Apr 25, 2023
Comment on lines
130
to
148
| template <typename GraphViewType, typename T> | ||
| auto create_constant_edge_property(raft::handle_t const& handle, | ||
| GraphViewType const& graph_view, | ||
| T constant_value) | ||
| { | ||
| edge_property_t<GraphViewType, T> edge_property(handle, graph_view); | ||
|
|
||
| auto mutable_view = edge_property.mutable_view(); | ||
|
|
||
| std::for_each( | ||
| thrust::make_zip_iterator(mutable_view.value_firsts().begin(), | ||
| mutable_view.edge_counts().begin()), | ||
| thrust::make_zip_iterator(mutable_view.value_firsts().end(), mutable_view.edge_counts().end()), | ||
| [&handle, constant_value](auto tuple) { | ||
| detail::scalar_fill(handle, thrust::get<0>(tuple), thrust::get<1>(tuple), constant_value); | ||
| }); | ||
|
|
||
| return edge_property; | ||
| } |
There was a problem hiding this comment.
We have fill_edge_src|dst_property() (https://github.com/rapidsai/cugraph/blob/branch-23.06/cpp/src/prims/fill_edge_src_dst_property.cuh).
So, it will be more consistent to create fill_edge_property() and call this instead of adding this function. I can quickly write fill_edge_property() if you agree.
There was a problem hiding this comment.
Merged your PR and then my change to use it.
seunghwak
approved these changes
Apr 27, 2023
| @@ -16,6 +16,7 @@ | |||
|
|
|||
| #pragma once | |||
|
|
|||
| #include <cugraph/detail/utility_wrappers.hpp> | |||
There was a problem hiding this comment.
Is this include still necessary?
|
/merge |
rapids-bot bot
pushed a commit
that referenced
this pull request
May 9, 2023
…ghts Separately (#3495) Closes #3486 Merge after #3513 - Merged Authors: - Alex Barghi (https://github.com/alexbarghi-nv) - Chuck Hastings (https://github.com/ChuckHastings) Approvers: - Rick Ratzel (https://github.com/rlratzel) - Seunghwa Kang (https://github.com/seunghwak) - Ray Douglass (https://github.com/raydouglass) URL: #3495
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Closes #3499
Some of our C++ algorithms require a graph to be weighted. The python layer will add "fake" edge weights of 1 to every edge in order to call these algorithms. This adds a persisted constant value for every edge that will continue to exist for the life span of the python graph object, even though it is only necessary for a handful of algorithms.
This PR addresses these concerns.
Many of the C++ algorithms already actually support the graph not having weights. A few should support it and had bugs that were corrected (induced subgraph, k-core, egonet). A few can easily support it but did not (legacy spectral algorithms).
The Louvain and Leiden algorithms need to have a weighted graph because they will coarsen the weights as the graph is coarsened, and the weights will eventually not be one while the algorithm is running. These require a slightly different approach.
This PR also added C API tests for unweighted graphs for these algorithms.