[Debian-med-packaging] Bug#1079676: imbalanced-learn: tests fail with scipy 1.14
Drew Parsons
dparsons at debian.org
Mon Aug 26 08:49:28 BST 2024
Source: imbalanced-learn
Version: 0.12.2-1
Severity: normal
imbalanced-learn is failing tests with scipy 1.14 (from experimental)
due to an API change
159s ␛[31m␛[1m___________ test_estimators_imblearn[AllKNN()-check_samplers_sparse] ___________␛[0m
159s
159s estimator = AllKNN()
159s check = functools.partial(<function check_samplers_sparse at 0x7f29b0afade0>, 'AllKNN')
159s request = <FixtureRequest for <Function test_estimators_imblearn[AllKNN()-check_samplers_sparse]>>
159s
159s @parametrize_with_checks(list(_tested_estimators()))
159s def test_estimators_imblearn(estimator, check, request):
159s # Common tests for estimator instances
159s with ignore_warnings(
159s category=(
159s FutureWarning,
159s ConvergenceWarning,
159s UserWarning,
159s FutureWarning,
159s )
159s ):
159s _set_checking_parameters(estimator)
159s > check(estimator)
159s
159s ␛[1m␛[31mimblearn/tests/test_common.py␛[0m:71:
159s _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
159s
159s name = 'AllKNN', sampler_orig = AllKNN()
159s
159s def check_samplers_sparse(name, sampler_orig):
159s sampler = clone(sampler_orig)
159s # check that sparse matrices can be passed through the sampler leading to
159s # the same results than dense
159s X, y = sample_dataset_generator()
159s X_sparse = sparse.csr_matrix(X)
159s X_res_sparse, y_res_sparse = sampler.fit_resample(X_sparse, y)
159s sampler = clone(sampler)
159s X_res, y_res = sampler.fit_resample(X, y)
159s assert sparse.issparse(X_res_sparse)
159s > assert_allclose(X_res_sparse.A, X_res, rtol=1e-5)
159s ␛[1m␛[31mE AttributeError: 'csr_matrix' object has no attribute 'A'␛[0m
159s
159s ␛[1m␛[31mimblearn/utils/estimator_checks.py␛[0m:312: AttributeError
test log at
https://ci.debian.net/packages/i/imbalanced-learn/unstable/amd64/50977859/
More information about the Debian-med-packaging
mailing list