Skip to content

Fix _estimate_mi discrete_features str and value check #13497

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

Merged
merged 10 commits into from
Apr 1, 2019
+21 −11
Merged

Fix _estimate_mi discrete_features str and value check #13497

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
a18cb3d
discrete_features str and value check
hermidalc Mar 23, 2019
eb3f556
Update if logic
hermidalc Mar 26, 2019
835bea8
Add discrete_features bad str value test
hermidalc Mar 26, 2019
2e78921
Remove unnecessary nested isinstance str check
hermidalc Mar 26, 2019
7474d6d
Add back nested isinstance str check
hermidalc Mar 26, 2019
954d260
New/updates to tests
hermidalc Mar 28, 2019
e67fb52
Merge remote-tracking branch 'upstream/master' into mi_discrete_featu…
hermidalc Mar 29, 2019
583a194
Add v0.21 whats new entry
hermidalc Mar 31, 2019
1a06ff2
Undo v0.21 whats new entry
hermidalc Apr 1, 2019
433fa02
Merge remote-tracking branch 'upstream/master' into mi_discrete_featu…
hermidalc Apr 1, 2019
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
14 changes: 8 additions & 6 deletions sklearn/feature_selection/mutual_info_.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@
from ..preprocessing import scale
from ..utils import check_random_state
from ..utils.fixes import _astype_copy_false
from ..utils.validation import check_X_y
from ..utils.validation import check_array, check_X_y
from ..utils.multiclass import check_classification_targets


Expand Down Expand Up @@ -247,14 +247,16 @@ def _estimate_mi(X, y, discrete_features='auto', discrete_target=False,
X, y = check_X_y(X, y, accept_sparse='csc', y_numeric=not discrete_target)
n_samples, n_features = X.shape

if discrete_features == 'auto':
discrete_features = issparse(X)

if isinstance(discrete_features, bool):
if isinstance(discrete_features, (str, bool)):
if isinstance(discrete_features, str):
if discrete_features == 'auto':
discrete_features = issparse(X)
else:
raise ValueError("Invalid string value for discrete_features.")
discrete_mask = np.empty(n_features, dtype=bool)
discrete_mask.fill(discrete_features)
else:
discrete_features = np.asarray(discrete_features)
discrete_features = check_array(discrete_features, ensure_2d=False)
if discrete_features.dtype != 'bool':
discrete_mask = np.zeros(n_features, dtype=bool)
discrete_mask[discrete_features] = True
Expand Down
18 changes: 13 additions & 5 deletions sklearn/feature_selection/tests/test_mutual_info.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -183,18 +183,26 @@ def test_mutual_info_options():
X_csr = csr_matrix(X)

for mutual_info in (mutual_info_regression, mutual_info_classif):
assert_raises(ValueError, mutual_info_regression, X_csr, y,
assert_raises(ValueError, mutual_info, X_csr, y,
discrete_features=False)
assert_raises(ValueError, mutual_info, X, y,
discrete_features='manual')
assert_raises(ValueError, mutual_info, X_csr, y,
discrete_features=[True, False, True])
assert_raises(IndexError, mutual_info, X, y,
discrete_features=[True, False, True, False])
assert_raises(IndexError, mutual_info, X, y, discrete_features=[1, 4])

mi_1 = mutual_info(X, y, discrete_features='auto', random_state=0)
mi_2 = mutual_info(X, y, discrete_features=False, random_state=0)

mi_3 = mutual_info(X_csr, y, discrete_features='auto',
random_state=0)
mi_4 = mutual_info(X_csr, y, discrete_features=True,
mi_3 = mutual_info(X_csr, y, discrete_features='auto', random_state=0)
mi_4 = mutual_info(X_csr, y, discrete_features=True, random_state=0)
mi_5 = mutual_info(X, y, discrete_features=[True, False, True],
random_state=0)
mi_6 = mutual_info(X, y, discrete_features=[0, 2], random_state=0)

assert_array_equal(mi_1, mi_2)
assert_array_equal(mi_3, mi_4)
assert_array_equal(mi_5, mi_6)

assert not np.allclose(mi_1, mi_3)