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.

Sign up for GitHub

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

nx-cugraph: add dijkstra sssp functions #4538

Merged
merged 5 commits into from
Jul 30, 2024
Merged

nx-cugraph: add dijkstra sssp functions #4538

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
72c32cd
nx-cugraph: add dijkstra sssp functions
eriknw Jul 15, 2024
42dd685
oops fix `can_run` signatures
eriknw Jul 16, 2024
cc84d5f
Merge branch 'branch-24.08' into sssp_dijkstra
eriknw Jul 21, 2024
342ad63
Merge branch 'branch-24.08' into sssp_dijkstra
rlratzel Jul 24, 2024
3d6a893
Merge branch 'branch-24.08' into sssp_dijkstra
nv-rliu Jul 29, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 9 additions & 1 deletion python/nx-cugraph/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -173,11 +173,19 @@ Below is the list of algorithms that are currently supported in nx-cugraph.
└─ <a href="https://networkx.org/documentation/stable/reference/algorithms/shortest_paths.html#module-networkx.algorithms.shortest_paths.weighted">weighted</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.all_pairs_bellman_ford_path.html#networkx.algorithms.shortest_paths.weighted.all_pairs_bellman_ford_path">all_pairs_bellman_ford_path</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.all_pairs_bellman_ford_path_length.html#networkx.algorithms.shortest_paths.weighted.all_pairs_bellman_ford_path_length">all_pairs_bellman_ford_path_length</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra.html#networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra">all_pairs_dijkstra</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra_path.html#networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra_path">all_pairs_dijkstra_path</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra_path_length.html#networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra_path_length">all_pairs_dijkstra_path_length</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.bellman_ford_path.html#networkx.algorithms.shortest_paths.weighted.bellman_ford_path">bellman_ford_path</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.bellman_ford_path_length.html#networkx.algorithms.shortest_paths.weighted.bellman_ford_path_length">bellman_ford_path_length</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.dijkstra_path.html#networkx.algorithms.shortest_paths.weighted.dijkstra_path">dijkstra_path</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.dijkstra_path_length.html#networkx.algorithms.shortest_paths.weighted.dijkstra_path_length">dijkstra_path_length</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford.html#networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford">single_source_bellman_ford</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path.html#networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path">single_source_bellman_ford_path</a>
└─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path_length.html#networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path_length">single_source_bellman_ford_path_length</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path_length.html#networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path_length">single_source_bellman_ford_path_length</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_dijkstra.html#networkx.algorithms.shortest_paths.weighted.single_source_dijkstra">single_source_dijkstra</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_dijkstra_path.html#networkx.algorithms.shortest_paths.weighted.single_source_dijkstra_path">single_source_dijkstra_path</a>
└─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.shortest_paths.weighted.single_source_dijkstra_path_length.html#networkx.algorithms.shortest_paths.weighted.single_source_dijkstra_path_length">single_source_dijkstra_path_length</a>
<a href="https://networkx.org/documentation/stable/reference/algorithms/traversal.html">traversal</a>
└─ <a href="https://networkx.org/documentation/stable/reference/algorithms/traversal.html#module-networkx.algorithms.traversal.breadth_first_search">breadth_first_search</a>
├─ <a href="https://networkx.org/documentation/stable/reference/algorithms/generated/networkx.algorithms.traversal.breadth_first_search.bfs_edges.html#networkx.algorithms.traversal.breadth_first_search.bfs_edges">bfs_edges</a>
Expand Down
36 changes: 34 additions & 2 deletions python/nx-cugraph/_nx_cugraph/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,6 +42,9 @@
# BEGIN: functions
"all_pairs_bellman_ford_path",
"all_pairs_bellman_ford_path_length",
"all_pairs_dijkstra",
"all_pairs_dijkstra_path",
"all_pairs_dijkstra_path_length",
"all_pairs_shortest_path",
"all_pairs_shortest_path_length",
"ancestors",
Expand Down Expand Up @@ -75,6 +78,8 @@
"descendants",
"descendants_at_distance",
"diamond_graph",
"dijkstra_path",
"dijkstra_path_length",
"dodecahedral_graph",
"edge_betweenness_centrality",
"ego_graph",
Expand Down Expand Up @@ -131,6 +136,9 @@
"single_source_bellman_ford",
"single_source_bellman_ford_path",
"single_source_bellman_ford_path_length",
"single_source_dijkstra",
"single_source_dijkstra_path",
"single_source_dijkstra_path_length",
"single_source_shortest_path",
"single_source_shortest_path_length",
"single_target_shortest_path",
Expand Down Expand Up @@ -171,8 +179,8 @@
"katz_centrality": "`nstart` isn't used (but is checked), and `normalized=False` is not supported.",
"louvain_communities": "`seed` parameter is currently ignored, and self-loops are not yet supported.",
"pagerank": "`dangling` parameter is not supported, but it is checked for validity.",
"shortest_path": "Negative weights are not yet supported, and method is ununsed.",
"shortest_path_length": "Negative weights are not yet supported, and method is ununsed.",
"shortest_path": "Negative weights are not yet supported.",
"shortest_path_length": "Negative weights are not yet supported.",
"single_source_bellman_ford": "Negative cycles are not yet supported. ``NotImplementedError`` will be raised if there are negative edge weights. We plan to support negative edge weights soon. Also, callable ``weight`` argument is not supported.",
"single_source_bellman_ford_path": "Negative cycles are not yet supported. ``NotImplementedError`` will be raised if there are negative edge weights. We plan to support negative edge weights soon. Also, callable ``weight`` argument is not supported.",
"single_source_bellman_ford_path_length": "Negative cycles are not yet supported. ``NotImplementedError`` will be raised if there are negative edge weights. We plan to support negative edge weights soon. Also, callable ``weight`` argument is not supported.",
Expand All @@ -187,12 +195,27 @@
"all_pairs_bellman_ford_path_length": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"all_pairs_dijkstra": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"all_pairs_dijkstra_path": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"all_pairs_dijkstra_path_length": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"bellman_ford_path": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"bellman_ford_path_length": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"dijkstra_path": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"dijkstra_path_length": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"ego_graph": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
Expand Down Expand Up @@ -227,6 +250,15 @@
"single_source_bellman_ford_path_length": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"single_source_dijkstra": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"single_source_dijkstra_path": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
"single_source_dijkstra_path_length": {
"dtype : dtype or None, optional": "The data type (np.float32, np.float64, or None) to use for the edge weights in the algorithm. If None, then dtype is determined by the edge values.",
},
# END: additional_parameters
},
}
Expand Down
54 changes: 30 additions & 24 deletions python/nx-cugraph/nx_cugraph/algorithms/shortest_paths/generic.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -43,7 +43,7 @@ def has_path(G, source, target):
def shortest_path(
G, source=None, target=None, weight=None, method="dijkstra", *, dtype=None
):
"""Negative weights are not yet supported, and method is ununsed."""
"""Negative weights are not yet supported."""
if method not in {"dijkstra", "bellman-ford"}:
raise ValueError(f"method not supported: {method}")
if weight is None:
Expand All @@ -53,9 +53,9 @@ def shortest_path(
# All pairs
if method == "unweighted":
paths = nxcg.all_pairs_shortest_path(G)
else:
# method == "dijkstra":
# method == 'bellman-ford':
elif method == "dijkstra":
paths = nxcg.all_pairs_dijkstra_path(G, weight=weight, dtype=dtype)
else: # method == 'bellman-ford':
paths = nxcg.all_pairs_bellman_ford_path(G, weight=weight, dtype=dtype)
if nx.__version__[:3] <= "3.4":
paths = dict(paths)
Expand All @@ -75,9 +75,11 @@ def shortest_path(
# From source
if method == "unweighted":
paths = nxcg.single_source_shortest_path(G, source)
else:
# method == "dijkstra":
# method == 'bellman-ford':
elif method == "dijkstra":
paths = nxcg.single_source_dijkstra_path(
G, source, weight=weight, dtype=dtype
)
else: # method == 'bellman-ford':
paths = nxcg.single_source_bellman_ford_path(
G, source, weight=weight, dtype=dtype
)
Expand Down Expand Up @@ -106,7 +108,7 @@ def _(G, source=None, target=None, weight=None, method="dijkstra", *, dtype=None
def shortest_path_length(
G, source=None, target=None, weight=None, method="dijkstra", *, dtype=None
):
"""Negative weights are not yet supported, and method is ununsed."""
"""Negative weights are not yet supported."""
if method not in {"dijkstra", "bellman-ford"}:
raise ValueError(f"method not supported: {method}")
if weight is None:
Expand All @@ -116,9 +118,11 @@ def shortest_path_length(
# All pairs
if method == "unweighted":
lengths = nxcg.all_pairs_shortest_path_length(G)
else:
# method == "dijkstra":
# method == 'bellman-ford':
elif method == "dijkstra":
lengths = nxcg.all_pairs_dijkstra_path_length(
G, weight=weight, dtype=dtype
)
else: # method == 'bellman-ford':
lengths = nxcg.all_pairs_bellman_ford_path_length(
G, weight=weight, dtype=dtype
)
Expand All @@ -127,31 +131,33 @@ def shortest_path_length(
lengths = nxcg.single_target_shortest_path_length(G, target)
if nx.__version__[:3] <= "3.4":
lengths = dict(lengths)
else:
# method == "dijkstra":
# method == 'bellman-ford':
elif method == "dijkstra":
lengths = nxcg.single_source_dijkstra_path_length(
G, target, weight=weight, dtype=dtype
)
else: # method == 'bellman-ford':
lengths = nxcg.single_source_bellman_ford_path_length(
G, target, weight=weight, dtype=dtype
)
elif target is None:
# From source
if method == "unweighted":
lengths = nxcg.single_source_shortest_path_length(G, source)
else:
# method == "dijkstra":
# method == 'bellman-ford':
lengths = dict(
nxcg.single_source_bellman_ford_path_length(
G, source, weight=weight, dtype=dtype
)
elif method == "dijkstra":
lengths = nxcg.single_source_dijkstra_path_length(
G, source, weight=weight, dtype=dtype
)
else: # method == 'bellman-ford':
lengths = nxcg.single_source_bellman_ford_path_length(
G, source, weight=weight, dtype=dtype
)
# From source to target
elif method == "unweighted":
G = _to_graph(G)
lengths = _bfs(G, source, None, "Source", return_type="length", target=target)
else:
# method == "dijkstra":
# method == 'bellman-ford':
elif method == "dijkstra":
lengths = nxcg.dijkstra_path_length(G, source, target, weight, dtype=dtype)
else: # method == 'bellman-ford':
lengths = nxcg.bellman_ford_path_length(G, source, target, weight, dtype=dtype)
return lengths

Expand Down
9 changes: 7 additions & 2 deletions python/nx-cugraph/nx_cugraph/algorithms/shortest_paths/unweighted.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,7 +61,12 @@ def bidirectional_shortest_path(G, source, target):
# TODO PERF: do bidirectional traversal in core
G = _to_graph(G)
if source not in G or target not in G:
raise nx.NodeNotFound(f"Either source {source} or target {target} is not in G")
if nx.__version__[:3] <= "3.3":
raise nx.NodeNotFound(
f"Either source {source} or target {target} is not in G"
)
missing = f"Source {source}" if source not in G else f"Target {target}"
raise nx.NodeNotFound(f"{missing} is not in G")
return _bfs(G, source, None, "Source", return_type="path", target=target)


Expand Down Expand Up @@ -131,7 +136,7 @@ def _bfs(
# return_type == "length-path"
return {source: 0}, {source: [source]}

if cutoff is None:
if cutoff is None or np.isinf(cutoff):
cutoff = -1
src_index = source if G.key_to_id is None else G.key_to_id[source]
distances, predecessors, node_ids = plc.bfs(
Expand Down
Loading
Loading