FIX LabelPropagation handling of sparce matrices #17085 #17384
There are no files selected for viewing
This file contains hidden or 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 hidden or 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 |
|---|---|---|
| @@ -1,9 +1,11 @@ | ||
| """ test the label propagation module """ | ||
|
|
||
| import pytest | ||
| import numpy as np | ||
| # Some tests fail for dok_matrix. | ||
| from scipy.sparse import (bsr_matrix, coo_matrix, csc_matrix, csr_matrix, | ||
| lil_matrix, issparse) | ||
|
|
||
| from sklearn.utils._testing import assert_warns | ||
| from sklearn.utils._testing import assert_no_warnings | ||
| from sklearn.semi_supervised import _label_propagation as label_propagation | ||
|
|
@@ -15,6 +17,9 @@ | |
| from numpy.testing import assert_array_almost_equal | ||
| from numpy.testing import assert_array_equal | ||
|
|
||
| SPARSE_TYPES = (bsr_matrix, coo_matrix, csc_matrix, csr_matrix, lil_matrix) | ||
| SPARSE_OR_DENSE = SPARSE_TYPES + (np.asarray,) | ||
|
|
||
| ESTIMATORS = [ | ||
| (label_propagation.LabelPropagation, {'kernel': 'rbf'}), | ||
| (label_propagation.LabelPropagation, {'kernel': 'knn', 'n_neighbors': 2}), | ||
|
|
@@ -73,18 +78,19 @@ def test_label_spreading_closed_form(): | |
| X, y = make_classification(n_classes=n_classes, n_samples=200, | ||
| random_state=0) | ||
| y[::3] = -1 | ||
| for sparse_or_dense in SPARSE_OR_DENSE: | ||
|
There was a problem hiding this comment. could you please parameterize the test for this one and There was a problem hiding this comment. @adrinjalali: I didn't write the original tests, so I'm probably not the right one to ask for improving these with regard to parametrization. |
||
| clf = label_propagation.LabelSpreading().fit(sparse_or_dense(X), y) | ||
| # adopting notation from Zhou et al (2004): | ||
| S = clf._build_graph() | ||
| Y = np.zeros((len(y), n_classes + 1)) | ||
| Y[np.arange(len(y)), y] = 1 | ||
| Y = Y[:, :-1] | ||
| for alpha in [0.1, 0.3, 0.5, 0.7, 0.9]: | ||
| expected = np.dot(np.linalg.inv(np.eye(len(S)) - alpha * S), Y) | ||
| expected /= expected.sum(axis=1)[:, np.newaxis] | ||
| clf = label_propagation.LabelSpreading(max_iter=10000, alpha=alpha) | ||
| clf.fit(sparse_or_dense(X), y) | ||
| assert_array_almost_equal(expected, clf.label_distributions_, 4) | ||
|
|
||
|
|
||
| def test_label_propagation_closed_form(): | ||
|
|
@@ -97,6 +103,7 @@ def test_label_propagation_closed_form(): | |
| unlabelled_idx = Y[:, (-1,)].nonzero()[0] | ||
| labelled_idx = (Y[:, (-1,)] == 0).nonzero()[0] | ||
|
|
||
| # This test fails for sparse matrices! | ||
| clf = label_propagation.LabelPropagation(max_iter=10000, | ||
| gamma=0.1) | ||
| clf.fit(X, y) | ||
|
|
@@ -121,40 +128,44 @@ def test_valid_alpha(): | |
| n_classes = 2 | ||
| X, y = make_classification(n_classes=n_classes, n_samples=200, | ||
| random_state=0) | ||
| for sparse_or_dense in SPARSE_OR_DENSE: | ||
| for alpha in [-0.1, 0, 1, 1.1, None]: | ||
| with pytest.raises(ValueError): | ||
| label_propagation.LabelSpreading(alpha=alpha).fit( | ||
| sparse_or_dense(X), y) | ||
|
|
||
|
|
||
| def test_convergence_speed(): | ||
| # This is a non-regression test for #5774 | ||
| X = np.array([[1., 0.], [0., 1.], [1., 2.5]]) | ||
| y = np.array([0, 1, -1]) | ||
| for sparse_or_dense in SPARSE_OR_DENSE: | ||
| mdl = label_propagation.LabelSpreading(kernel='rbf', max_iter=5000) | ||
| mdl.fit(sparse_or_dense(X), y) | ||
|
|
||
| # this should converge quickly: | ||
| assert mdl.n_iter_ < 10 | ||
| assert_array_equal(mdl.predict(X), [0, 1, 1]) | ||
|
|
||
|
|
||
| def test_convergence_warning(): | ||
| # This is a non-regression test for #5774 | ||
| X = np.array([[1., 0.], [0., 1.], [1., 2.5]]) | ||
| y = np.array([0, 1, -1]) | ||
| for sparse_or_dense in SPARSE_OR_DENSE: | ||
| mdl = label_propagation.LabelSpreading(kernel='rbf', max_iter=1) | ||
| assert_warns(ConvergenceWarning, mdl.fit, sparse_or_dense(X), y) | ||
| assert mdl.n_iter_ == mdl.max_iter | ||
|
|
||
| mdl = label_propagation.LabelPropagation(kernel='rbf', max_iter=1) | ||
| assert_warns(ConvergenceWarning, mdl.fit, sparse_or_dense(X), y) | ||
| assert mdl.n_iter_ == mdl.max_iter | ||
|
|
||
| mdl = label_propagation.LabelSpreading(kernel='rbf', max_iter=500) | ||
| assert_no_warnings(mdl.fit, sparse_or_dense(X), y) | ||
|
|
||
| mdl = label_propagation.LabelPropagation(kernel='rbf', max_iter=500) | ||
| assert_no_warnings(mdl.fit, sparse_or_dense(X), y) | ||
|
|
||
|
|
||
| def test_label_propagation_non_zero_normalizer(): | ||
|
|
@@ -163,10 +174,11 @@ def test_label_propagation_non_zero_normalizer(): | |
| # https://github.com/scikit-learn/scikit-learn/pull/15946 | ||
| X = np.array([[100., 100.], [100., 100.], [0., 0.], [0., 0.]]) | ||
| y = np.array([0, 1, -1, -1]) | ||
| for sparse_or_dense in SPARSE_OR_DENSE: | ||
| mdl = label_propagation.LabelSpreading(kernel='knn', | ||
| max_iter=100, | ||
| n_neighbors=1) | ||
| assert_no_warnings(mdl.fit, sparse_or_dense(X), y) | ||
|
|
||
|
|
||
| def test_predict_sparse_callable_kernel(): | ||
|
|
@@ -196,10 +208,11 @@ def topk_rbf(X, Y=None, n_neighbors=10, gamma=1e-5): | |
| test_size=n_test, | ||
| random_state=0) | ||
|
|
||
| for sparse_or_dense in SPARSE_OR_DENSE: | ||
| model = label_propagation.LabelSpreading(kernel=topk_rbf) | ||
| model.fit(sparse_or_dense(X_train), y_train) | ||
| assert model.score(X_test, y_test) >= 0.9 | ||
|
|
||
| model = label_propagation.LabelPropagation(kernel=topk_rbf) | ||
| model.fit(sparse_or_dense(X_train), y_train) | ||
| assert model.score(X_test, y_test) >= 0.9 |
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.
should this use
_validate_data? ping @NicolasHugThere 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'm also really not sure why this fixes the issue
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.
We only use it in
fitfor now, socheck_arrayis fine hereThere 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.
Thanks for the feedback @adrinjalali: PR #17085 already contains tests extended to the sparse case. Could you please advise a newbie regarding regression tests?