-
Notifications
You must be signed in to change notification settings - Fork 300
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
nx-cugraph: add CC for undirected graphs to fix k-truss #3965
Merged
Merged
Changes from all commits
Commits
Show all changes
5 commits
Select commit
Hold shift + click to select a range
447cdfd
nx-cugraph: add CC for undirected graphs to fix k-truss
eriknw c5350c2
Add note to k-truss and skip very slow tests
eriknw a056b57
Merge branch 'branch-23.12' into cc_fix_ktruss
eriknw 137bad9
Test k_truss
eriknw dd75489
Merge branch 'branch-23.12' into cc_fix_ktruss
eriknw File filter
Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
There are no files selected for viewing
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
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
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
13 changes: 13 additions & 0 deletions
13
python/nx-cugraph/nx_cugraph/algorithms/components/__init__.py
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
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,13 @@ | ||
# Copyright (c) 2023, NVIDIA CORPORATION. | ||
# Licensed under the Apache License, Version 2.0 (the "License"); | ||
# you may not use this file except in compliance with the License. | ||
# You may obtain a copy of the License at | ||
# | ||
# http://www.apache.org/licenses/LICENSE-2.0 | ||
# | ||
# Unless required by applicable law or agreed to in writing, software | ||
# distributed under the License is distributed on an "AS IS" BASIS, | ||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
# See the License for the specific language governing permissions and | ||
# limitations under the License. | ||
from .connected import * |
130 changes: 130 additions & 0 deletions
130
python/nx-cugraph/nx_cugraph/algorithms/components/connected.py
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
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,130 @@ | ||
# Copyright (c) 2023, NVIDIA CORPORATION. | ||
# Licensed under the Apache License, Version 2.0 (the "License"); | ||
# you may not use this file except in compliance with the License. | ||
# You may obtain a copy of the License at | ||
# | ||
# http://www.apache.org/licenses/LICENSE-2.0 | ||
# | ||
# Unless required by applicable law or agreed to in writing, software | ||
# distributed under the License is distributed on an "AS IS" BASIS, | ||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
# See the License for the specific language governing permissions and | ||
# limitations under the License. | ||
import itertools | ||
|
||
import cupy as cp | ||
import networkx as nx | ||
import pylibcugraph as plc | ||
|
||
from nx_cugraph.convert import _to_undirected_graph | ||
from nx_cugraph.utils import _groupby, networkx_algorithm, not_implemented_for | ||
|
||
from ..isolate import _isolates | ||
|
||
__all__ = [ | ||
"number_connected_components", | ||
"connected_components", | ||
"is_connected", | ||
"node_connected_component", | ||
] | ||
|
||
|
||
@not_implemented_for("directed") | ||
@networkx_algorithm | ||
def number_connected_components(G): | ||
return sum(1 for _ in connected_components(G)) | ||
# PREFERRED IMPLEMENTATION, BUT PLC DOES NOT HANDLE ISOLATED VERTICES WELL | ||
# G = _to_undirected_graph(G) | ||
# unused_node_ids, labels = plc.weakly_connected_components( | ||
# resource_handle=plc.ResourceHandle(), | ||
# graph=G._get_plc_graph(), | ||
# offsets=None, | ||
# indices=None, | ||
# weights=None, | ||
# labels=None, | ||
# do_expensive_check=False, | ||
# ) | ||
# return cp.unique(labels).size | ||
|
||
|
||
@number_connected_components._can_run | ||
def _(G): | ||
# NetworkX <= 3.2.1 does not check directedness for us | ||
try: | ||
return not G.is_directed() | ||
except Exception: | ||
return False | ||
|
||
|
||
@not_implemented_for("directed") | ||
@networkx_algorithm | ||
def connected_components(G): | ||
G = _to_undirected_graph(G) | ||
if G.src_indices.size == 0: | ||
# TODO: PLC doesn't handle empty graphs (or isolated nodes) gracefully! | ||
return [{key} for key in G._nodeiter_to_iter(range(len(G)))] | ||
node_ids, labels = plc.weakly_connected_components( | ||
resource_handle=plc.ResourceHandle(), | ||
graph=G._get_plc_graph(), | ||
offsets=None, | ||
indices=None, | ||
weights=None, | ||
labels=None, | ||
do_expensive_check=False, | ||
) | ||
groups = _groupby(labels, node_ids) | ||
it = (G._nodearray_to_set(connected_ids) for connected_ids in groups.values()) | ||
# TODO: PLC doesn't handle isolated vertices yet, so this is a temporary fix | ||
isolates = _isolates(G) | ||
if isolates.size > 0: | ||
isolates = isolates[isolates > node_ids.max()] | ||
if isolates.size > 0: | ||
it = itertools.chain( | ||
it, ({node} for node in G._nodearray_to_list(isolates)) | ||
) | ||
return it | ||
|
||
|
||
@not_implemented_for("directed") | ||
@networkx_algorithm | ||
def is_connected(G): | ||
G = _to_undirected_graph(G) | ||
if len(G) == 0: | ||
raise nx.NetworkXPointlessConcept( | ||
"Connectivity is undefined for the null graph." | ||
) | ||
for community in connected_components(G): | ||
return len(community) == len(G) | ||
raise RuntimeError # pragma: no cover | ||
# PREFERRED IMPLEMENTATION, BUT PLC DOES NOT HANDLE ISOLATED VERTICES WELL | ||
# unused_node_ids, labels = plc.weakly_connected_components( | ||
# resource_handle=plc.ResourceHandle(), | ||
# graph=G._get_plc_graph(), | ||
# offsets=None, | ||
# indices=None, | ||
# weights=None, | ||
# labels=None, | ||
# do_expensive_check=False, | ||
# ) | ||
# return labels.size == len(G) and cp.unique(labels).size == 1 | ||
|
||
|
||
@not_implemented_for("directed") | ||
@networkx_algorithm | ||
def node_connected_component(G, n): | ||
# We could also do plain BFS from n | ||
G = _to_undirected_graph(G) | ||
node_id = n if G.key_to_id is None else G.key_to_id[n] | ||
node_ids, labels = plc.weakly_connected_components( | ||
resource_handle=plc.ResourceHandle(), | ||
graph=G._get_plc_graph(), | ||
offsets=None, | ||
indices=None, | ||
weights=None, | ||
labels=None, | ||
do_expensive_check=False, | ||
) | ||
indices = cp.nonzero(node_ids == node_id)[0] | ||
if indices.size == 0: | ||
return {n} | ||
return G._nodearray_to_set(node_ids[labels == labels[indices[0]]]) |
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
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
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
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
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,30 @@ | ||
# Copyright (c) 2023, NVIDIA CORPORATION. | ||
# Licensed under the Apache License, Version 2.0 (the "License"); | ||
# you may not use this file except in compliance with the License. | ||
# You may obtain a copy of the License at | ||
# | ||
# http://www.apache.org/licenses/LICENSE-2.0 | ||
# | ||
# Unless required by applicable law or agreed to in writing, software | ||
# distributed under the License is distributed on an "AS IS" BASIS, | ||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
# See the License for the specific language governing permissions and | ||
# limitations under the License. | ||
import networkx as nx | ||
import pytest | ||
|
||
import nx_cugraph as nxcg | ||
|
||
|
||
@pytest.mark.parametrize( | ||
"get_graph", [nx.florentine_families_graph, nx.les_miserables_graph] | ||
) | ||
def test_k_truss(get_graph): | ||
Gnx = get_graph() | ||
Gcg = nxcg.from_networkx(Gnx, preserve_all_attrs=True) | ||
for k in range(10): | ||
Hnx = nx.k_truss(Gnx, k) | ||
Hcg = nxcg.k_truss(Gcg, k) | ||
assert nx.utils.graphs_equal(Hnx, nxcg.to_networkx(Hcg)) | ||
if Hnx.number_of_edges() == 0: | ||
break |
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
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
Original file line number | Diff line number | Diff line change |
---|---|---|
|
@@ -21,40 +21,56 @@ | |
import cupy as cp | ||
import numpy as np | ||
|
||
try: | ||
from itertools import pairwise # Python >=3.10 | ||
except ImportError: | ||
|
||
def pairwise(it): | ||
it = iter(it) | ||
for prev in it: | ||
for cur in it: | ||
yield (prev, cur) | ||
prev = cur | ||
|
||
|
||
__all__ = ["index_dtype", "_groupby", "_seed_to_int", "_get_int_dtype"] | ||
|
||
# This may switch to np.uint32 at some point | ||
index_dtype = np.int32 | ||
|
||
|
||
def _groupby(groups: cp.ndarray, values: cp.ndarray) -> dict[int, cp.ndarray]: | ||
def _groupby( | ||
groups: cp.ndarray, values: cp.ndarray, groups_are_canonical: bool = False | ||
) -> dict[int, cp.ndarray]: | ||
Comment on lines
+42
to
+44
There was a problem hiding this comment. Choose a reason for hiding this commentThe reason will be displayed to describe this comment to others. Learn more. I'm really happy with this groupby implementation btw. Using |
||
"""Perform a groupby operation given an array of group IDs and array of values. | ||
|
||
Parameters | ||
---------- | ||
groups : cp.ndarray | ||
Array that holds the group IDs. | ||
Group IDs are assumed to be consecutive integers from 0. | ||
values : cp.ndarray | ||
Array of values to be grouped according to groups. | ||
Must be the same size as groups array. | ||
groups_are_canonical : bool, default False | ||
Whether the group IDs are consecutive integers beginning with 0. | ||
|
||
Returns | ||
------- | ||
dict with group IDs as keys and cp.ndarray as values. | ||
""" | ||
# It would actually be easy to support groups that aren't consecutive integers, | ||
# but let's wait until we need it to implement it. | ||
sorted_groups = cp.argsort(groups) | ||
sorted_values = values[sorted_groups] | ||
rv = {} | ||
start = 0 | ||
for i, end in enumerate( | ||
[*(cp.nonzero(cp.diff(groups[sorted_groups]))[0] + 1).tolist(), groups.size] | ||
): | ||
rv[i] = sorted_values[start:end] | ||
start = end | ||
return rv | ||
if groups.size == 0: | ||
return {} | ||
sort_indices = cp.argsort(groups) | ||
sorted_groups = groups[sort_indices] | ||
sorted_values = values[sort_indices] | ||
prepend = 1 if groups_are_canonical else sorted_groups[0] + 1 | ||
left_bounds = cp.nonzero(cp.diff(sorted_groups, prepend=prepend))[0] | ||
boundaries = pairwise(itertools.chain(left_bounds.tolist(), [groups.size])) | ||
if groups_are_canonical: | ||
it = enumerate(boundaries) | ||
else: | ||
it = zip(sorted_groups[left_bounds].tolist(), boundaries) | ||
return {group: sorted_values[start:end] for group, (start, end) in it} | ||
|
||
|
||
def _seed_to_int(seed: int | Random | None) -> int: | ||
|
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
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.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I fixed this upstream here: networkx/networkx#7074