-
Notifications
You must be signed in to change notification settings - Fork 344
Open
Labels
bugSomething isn't workingSomething isn't working
Description
Version
dev (26.02)
Which installation method(s) does this occur on?
Source
Describe the bug.
Tests in python/cugraph/cugraph/tests/sampling/test_egonet_mg.py use cugraph.batched_ego_graphs, which was removed in #5133. See also #4708.
I encountered this when running all cugraph tests locally on a multi-gpu system.
The easiest solution would be to remove the test file, but a better solution may be to update it to not use batched_ego_graphs. Also, I'm surprised I'm the first to catch this since #5133 was merged last July!
Minimum reproducible example
Run `test_egonet_mg.py` tests on a multi-gpu system.Relevant log output
_ ERROR at setup of test_dask_mg_ego_graphs[graph_file:/home/nfs/erwelch/.cugraph/datasets/karate.csv-directed:True-seeds:0-radius:1] _
input_combo = {'directed': True, 'graph_file': PosixPath('/home/nfs/erwelch/.cugraph/datasets/karate.csv'), 'radius': 1, 'seeds': 0}
@pytest.fixture(scope="module")
def input_expected_output(input_combo): """
This fixture returns the inputs and expected results from the egonet algo. (based on cuGraph batched_ego_graphs) which can be used for validation.
""" ····
input_data_path = input_combo["graph_file"] directed = input_combo["directed"]
seeds = input_combo["seeds"] radius = input_combo["radius"]
G = utils.generate_cugraph_graph_from_file( input_data_path, directed=directed, edgevals=True
) ····
> sg_cugraph_ego_graphs = cugraph.batched_ego_graphs(G, seeds=seeds, radius=radius) ^^^^^^^^^^^^^^^^^^^^^^^^^^
E AttributeError: module 'cugraph' has no attribute 'batched_ego_graphs'
tests/sampling/test_egonet_mg.py:73: AttributeErrorEnvironment details
Other/Misc.
No response
Code of Conduct
- I agree to follow cuGraph's Code of Conduct
- I have searched the open bugs and have found no duplicates for this bug report
Metadata
Metadata
Assignees
Labels
bugSomething isn't workingSomething isn't working