Your message dated Mon, 20 Jul 2020 20:15:10 +0200
with message-id <[email protected]>
and subject line Re: Bug#963822: numpy breaks scikit-learn autopkgtest:
test_set_estimator_none[drop] fails
has caused the Debian Bug report #963822,
regarding numpy breaks scikit-learn autopkgtest: test_set_estimator_none[drop]
fails
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
963822: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=963822
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Source: numpy, scikit-learn
Control: found -1 numpy/1:1.19.0-1
Control: found -1 scikit-learn/0.22.2.post1+dfsg-7
Severity: serious
Tags: sid bullseye
X-Debbugs-CC: [email protected]
User: [email protected]
Usertags: breaks needs-update
Dear maintainer(s),
With a recent upload of numpy the autopkgtest of scikit-learn fails in
testing when that autopkgtest is run with the binary packages of numpy
from unstable. It passes when run with only packages from testing. In
tabular form:
pass fail
numpy from testing 1:1.19.0-1
scikit-learn from testing 0.22.2.post1+dfsg-7
all others from testing from testing
I copied some of the output at the bottom of this report.
Currently this regression is blocking the migration of numpy to testing
[1]. Due to the nature of this issue, I filed this bug report against
both packages. Can you please investigate the situation and reassign the
bug to the right package?
More information about this bug and the reason for filing it can be found on
https://wiki.debian.org/ContinuousIntegration/RegressionEmailInformation
Paul
[1] https://qa.debian.org/excuses.php?package=numpy
https://ci.debian.net/data/autopkgtest/testing/amd64/s/scikit-learn/6058472/log.gz
=================================== FAILURES
===================================
________________________ test_set_estimator_none[drop]
_________________________
drop = 'drop'
@pytest.mark.parametrize("drop", [None, 'drop'])
def test_set_estimator_none(drop):
"""VotingClassifier set_params should be able to set estimators
as None or
drop"""
# Test predict
clf1 = LogisticRegression(random_state=123)
clf2 = RandomForestClassifier(n_estimators=10, random_state=123)
clf3 = GaussianNB()
eclf1 = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2),
('nb', clf3)],
voting='hard', weights=[1, 0,
0.5]).fit(X, y)
eclf2 = VotingClassifier(estimators=[('lr', clf1), ('rf', clf2),
('nb', clf3)],
voting='hard', weights=[1, 1, 0.5])
with pytest.warns(None) as record:
eclf2.set_params(rf=drop).fit(X, y)
> assert record if drop is None else not record
E assert False
/usr/lib/python3/dist-packages/sklearn/ensemble/tests/test_voting.py:378:
AssertionError
________________ test_logistic_regression_path_convergence_fail
________________
def test_logistic_regression_path_convergence_fail():
rng = np.random.RandomState(0)
X = np.concatenate((rng.randn(100, 2) + [1, 1], rng.randn(100, 2)))
y = [1] * 100 + [-1] * 100
Cs = [1e3]
# Check that the convergence message points to both a model agnostic
# advice (scaling the data) and to the logistic regression specific
# documentation that includes hints on the solver configuration.
with pytest.warns(ConvergenceWarning) as record:
_logistic_regression_path(
X, y, Cs=Cs, tol=0., max_iter=1, random_state=0, verbose=0)
> assert len(record) == 1
E assert 6 == 1
E -6
E +1
/usr/lib/python3/dist-packages/sklearn/linear_model/tests/test_logistic.py:401:
AssertionError
signature.asc
Description: OpenPGP digital signature
--- End Message ---
--- Begin Message ---
Version: 0.23.1-1
On 2020-06-27 22:12, Paul Gevers wrote:
> With a recent upload of numpy the autopkgtest of scikit-learn fails in
> testing when that autopkgtest is run with the binary packages of numpy
> from unstable.
> Currently this regression is blocking the migration of numpy to testing
> [1]. Due to the nature of this issue, I filed this bug report against
> both packages. Can you please investigate the situation and reassign the
> bug to the right package?
The issue actually originates in scipy, a dependency of scikit-learn. It
was fixed there in 1.5.0 AFAICT which has not yet been uploaded to
unstable, so I disabled the failing test for builds where scipy < 1.5.
I'll file an issue with scipy upstream.
--- End Message ---
