Closed eriknw closed 1 year ago
I'm also not sure if these are the best algorithms for connected component-related functions. Connected components are an important class of algorithms that we should add soon (for example, we have an implementation of FastSV in a Jupyter notebook that we can port over), so perhaps we'll revisit these algorithms then.
Patch coverage: 87.71
% and project coverage change: +2.46
:tada:
Comparison is base (
6de1fd6
) 69.16% compared to head (029a639
) 71.62%.
:mega: This organization is not using Codecov’s GitHub App Integration. We recommend you install it so Codecov can continue to function properly for your repositories. Learn more
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
Merging so we can do a release before PyConUS. May be able to benchmark and shakedown some algorithms later, but they're probably good enough.
BFS is a workhorse! There are still many BFS-based algorithms still to do.
My main TODO is to benchmark these to give them a proper shakedown. I like to shake down most algorithms we add. Also, I wonder whether we should use
A.T
orAT
in a couple places. I also wonder if we can come up with a heuristic to detect negative cycles.It may be nice to know whether directed graphs are structurally symmetric. If we knew this, we could improve
is_weakly_connected
and other algorithms.Also, there may be a couple NetworkX bugs to fix or behaviors to match.
Implemented:
is_connected
is_weakly_connected
node_connected_component
all_pairs_shortest_path_length
negative_edge_cycle
single_source_shortest_path_length
single_target_shortest_path_length
bfs_layers
descendants_at_distance