trafficstars

Extend the pylibcugraph similarity algorithms to python cugraph. MG implementations are now supported closes #2690

Sep 18 '22 15:09 jnke2016

@ChuckHastings Can you take a look at this PR when you get a chance? I just want to make sure graph_functions and the similarity algorithms (so far I only implemented jaccard_coefficients) are written according to the discussion in your PR (the equivalent C++ one). This PR builds successfully but doesn't run as I get a Not Implemented error at the python layer. Once I no longer get a Not Implemented, I will update pylibcugraph jaccard to extract the first and second from cugraph_vertex_pairs_t* and will return the trio and not just the similarity results.

Sep 18 '22 15:09 jnke2016

rerun tests

Nov 16 '22 13:11 jnke2016

rerun tests

Nov 16 '22 20:11 jnke2016

Codecov Report

Base: 60.10% // Head: 59.64% // Decreases project coverage by -0.46% :warning:

Coverage data is based on head (af4f380) compared to base (5e1baf1). Patch coverage: 50.82% of modified lines in pull request are covered.

Additional details and impacted files

@@               Coverage Diff                @@
##           branch-22.12    #2704      +/-   ##
================================================
- Coverage         60.10%   59.64%   -0.47%     
================================================
  Files               123      129       +6     
  Lines              7277     7632     +355     
================================================
+ Hits               4374     4552     +178     
- Misses             2903     3080     +177

Impacted Files	Coverage Δ
python/cugraph/cugraph/utilities/utils.py	`71.95% <ø> (ø)`
...ure/graph_implementation/simpleDistributedGraph.py	`14.64% <5.12%> (-1.56%)`	:arrow_down:
...on/cugraph/cugraph/dask/link_prediction/jaccard.py	`21.15% <21.15%> (ø)`
...on/cugraph/cugraph/dask/link_prediction/overlap.py	`21.15% <21.15%> (ø)`
...n/cugraph/cugraph/dask/link_prediction/sorensen.py	`21.15% <21.15%> (ø)`
...raph/structure/graph_implementation/simpleGraph.py	`78.28% <70.58%> (-0.63%)`	:arrow_down:
...ph/cugraph/experimental/link_prediction/overlap.py	`87.80% <87.80%> (ø)`
...h/cugraph/experimental/link_prediction/sorensen.py	`87.80% <87.80%> (ø)`
...ph/cugraph/experimental/link_prediction/jaccard.py	`92.68% <92.68%> (ø)`
python/cugraph/cugraph/community/leiden.py	`100.00% <100.00%> (ø)`
... and 9 more