Assert or ignore all sequence of sequences deprecation warnings

jnothman · ogrisel · commit 28eb1ebe15d6 · 2014-06-05T11:20:36.000+02:00
diff --git a/sklearn/datasets/tests/test_samples_generator.py b/sklearn/datasets/tests/test_samples_generator.py
@@ -13,6 +13,7 @@
 from sklearn.utils.testing import assert_true
 from sklearn.utils.testing import assert_less
 from sklearn.utils.testing import assert_raises
+from sklearn.utils.testing import assert_warns
 
 from sklearn.datasets import make_classification
 from sklearn.datasets import make_multilabel_classification
@@ -131,11 +132,11 @@ def test_make_classification_informative_features():
                   n_clusters_per_class=2)
 
 
-def test_make_multilabel_classification():
+def test_make_multilabel_classification_return_sequences():
     for allow_unlabeled, min_length in zip((True, False), (0, 1)):
-        X, Y = make_multilabel_classification(n_samples=100, n_features=20,
-                                              n_classes=3, random_state=0,
-                                              allow_unlabeled=allow_unlabeled)
+        X, Y = assert_warns(DeprecationWarning, make_multilabel_classification,
+                            n_samples=100, n_features=20, n_classes=3,
+                            random_state=0, allow_unlabeled=allow_unlabeled)
         assert_equal(X.shape, (100, 20), "X shape mismatch")
         if not allow_unlabeled:
             assert_equal(max([max(y) for y in Y]), 2)
diff --git a/sklearn/metrics/tests/test_metrics.py b/sklearn/metrics/tests/test_metrics.py
@@ -1682,7 +1682,7 @@ def test_multilabel_representation_invariance():
             metric.__module__ = 'tmp'
             metric.__name__ = name
         metric = ignore_warnings(metric)
-        measure = metric(y1, y2)
+        measure = assert_warns(DeprecationWarning, metric, y1, y2)
 
         # Check representation invariance
         assert_almost_equal(metric(y1_binary_indicator,
@@ -1894,7 +1894,8 @@ def test_normalize_option_multilabel_classification():
         metrics = ALL_METRICS[name]
 
         # List of list of labels
-        measure = metrics(y_true, y_pred, normalize=True)
+        measure = assert_warns(DeprecationWarning, metrics, y_true, y_pred,
+                               normalize=True)
         assert_greater(measure, 0,
                        msg="We failed to test correctly the normalize option")
         assert_almost_equal(metrics(y_true, y_pred, normalize=False)
diff --git a/sklearn/preprocessing/tests/test_label.py b/sklearn/preprocessing/tests/test_label.py
@@ -6,6 +6,8 @@
 from sklearn.utils.testing import assert_raises
 from sklearn.utils.testing import assert_true
 from sklearn.utils.testing import assert_false
+from sklearn.utils.testing import assert_warns
+from sklearn.utils.testing import ignore_warnings
 
 from sklearn.preprocessing.label import LabelBinarizer
 from sklearn.preprocessing.label import MultiLabelBinarizer
@@ -51,6 +53,7 @@ def test_label_binarizer():
     assert_array_equal(lb.inverse_transform(got), inp)
 
 
+@ignore_warnings
 def test_label_binarizer_column_y():
     # first for binary classification vs multi-label with 1 possible class
     # lists are multi-label, array is multi-class :-/
@@ -125,7 +128,7 @@ def test_label_binarizer_multilabel():
     indicator_mat = np.array([[0, 1, 1],
                               [1, 0, 0],
                               [1, 1, 0]])
-    got = lb.fit_transform(inp)
+    got = assert_warns(DeprecationWarning, lb.fit_transform, inp)
     assert_true(lb.multilabel_)
     assert_array_equal(indicator_mat, got)
     assert_equal(lb.inverse_transform(got), inp)
@@ -142,7 +145,7 @@ def test_label_binarizer_multilabel():
                          [1, 0],
                          [0, 1],
                          [1, 1]])
-    got = lb.fit_transform(inp)
+    got = assert_warns(DeprecationWarning, lb.fit_transform, inp)
     assert_true(lb.multilabel_)
     assert_array_equal(expected, got)
     assert_equal([set(x) for x in lb.inverse_transform(got)],
@@ -155,7 +158,7 @@ def test_label_binarizer_errors():
     lb = LabelBinarizer().fit(one_class)
     assert_false(lb.multilabel_)
 
-    multi_label = [(2, 3), (0,), (0, 2)]
+    multi_label = np.array([[0, 0, 1, 0], [1, 0, 1, 0]])
     assert_raises(ValueError, lb.transform, multi_label)
 
     lb = LabelBinarizer()
@@ -229,7 +232,8 @@ def test_label_binarizer_multilabel_unlabeled():
     Y = np.array([[1, 1],
                   [1, 0],
                   [0, 0]])
-    assert_array_equal(lb.fit_transform(y), Y)
+    assert_array_equal(assert_warns(DeprecationWarning,
+                                    lb.fit_transform, y), Y)
 
 
 def test_label_binarize_with_multilabel_indicator():
diff --git a/sklearn/tests/test_multiclass.py b/sklearn/tests/test_multiclass.py
@@ -7,6 +7,7 @@
 from sklearn.utils.testing import assert_true
 from sklearn.utils.testing import assert_false
 from sklearn.utils.testing import assert_raises
+from sklearn.utils.testing import assert_warns
 
 from sklearn.utils.testing import assert_greater
 from sklearn.multiclass import OneVsRestClassifier
@@ -86,7 +87,9 @@ def test_ovr_multilabel():
                      LinearRegression(), Ridge(),
                      ElasticNet(), Lasso(alpha=0.5)):
         # test input as lists of tuples
-        clf = OneVsRestClassifier(base_clf).fit(X, y)
+        clf = assert_warns(DeprecationWarning,
+                           OneVsRestClassifier(base_clf).fit,
+                           X, y)
         assert_equal(set(clf.classes_), classes)
         y_pred = clf.predict([[0, 4, 4]])[0]
         assert_equal(set(y_pred), set(["spam", "eggs"]))
@@ -115,6 +118,7 @@ def test_ovr_multilabel_dataset():
                                                        n_labels=2,
                                                        length=50,
                                                        allow_unlabeled=au,
+                                                       return_indicator=True,
                                                        random_state=0)
         X_train, Y_train = X[:80], Y[:80]
         X_test, Y_test = X[80:], Y[80:]
@@ -138,6 +142,7 @@ def test_ovr_multilabel_predict_proba():
                                                        n_labels=3,
                                                        length=50,
                                                        allow_unlabeled=au,
+                                                       return_indicator=True,
                                                        random_state=0)
         X_train, Y_train = X[:80], Y[:80]
         X_test, Y_test = X[80:], Y[80:]
@@ -157,8 +162,8 @@ def test_ovr_multilabel_predict_proba():
 
         # predict assigns a label if the probability that the
         # sample has the label is greater than 0.5.
-        pred = [tuple(l.nonzero()[0]) for l in (Y_proba > 0.5)]
-        assert_equal(pred, Y_pred)
+        pred = Y_proba > .5
+        assert_array_equal(pred, Y_pred)
 
 
 def test_ovr_single_label_predict_proba():
@@ -189,12 +194,13 @@ def test_ovr_multilabel_decision_function():
                                                    n_labels=3,
                                                    length=50,
                                                    allow_unlabeled=True,
+                                                   return_indicator=True,
                                                    random_state=0)
     X_train, Y_train = X[:80], Y[:80]
     X_test, Y_test = X[80:], Y[80:]
     clf = OneVsRestClassifier(svm.SVC()).fit(X_train, Y_train)
-    assert_array_equal((clf.decision_function(X_test) > 0).nonzero()[1],
-                       np.hstack(clf.predict(X_test)))
+    assert_array_equal((clf.decision_function(X_test) > 0).astype(int),
+                       clf.predict(X_test))
 
 
 def test_ovr_single_label_decision_function():
diff --git a/sklearn/utils/multiclass.py b/sklearn/utils/multiclass.py
@@ -68,11 +68,6 @@ def unique_labels(*ys):
     array([1, 2, 3, 4])
     >>> unique_labels([1, 2, 10], [5, 11])
     array([ 1,  2,  5, 10, 11])
-    >>> unique_labels(np.array([[0.0, 1.0], [1.0, 1.0]]), np.zeros((2, 2)))
-    array([0, 1])
-    >>> unique_labels([(1, 2), (3,)], [(1, 2), tuple()])
-    array([1, 2, 3])
-
     """
     if not ys:
         raise ValueError('No argument has been passed.')
@@ -152,6 +147,8 @@ def is_label_indicator_matrix(y):
 def is_sequence_of_sequences(y):
     """ Check if ``y`` is in the sequence of sequences format (multilabel).
 
+    This format is DEPRECATED.
+
     Parameters
     ----------
     y : sequence or array.
@@ -160,22 +157,6 @@ def is_sequence_of_sequences(y):
     -------
     out : bool,
         Return ``True``, if ``y`` is a sequence of sequences else ``False``.
-
-    >>> import numpy as np
-    >>> is_sequence_of_sequences([0, 1, 0, 1])
-    False
-    >>> is_sequence_of_sequences([[1], [0, 2], []])
-    True
-    >>> is_sequence_of_sequences(np.array([[1], [0, 2], []], dtype=object))
-    True
-    >>> is_sequence_of_sequences([(1,), (0, 2), ()])
-    True
-    >>> is_sequence_of_sequences(np.array([[1, 0], [0, 0]]))
-    False
-    >>> is_sequence_of_sequences(np.array([[1], [0], [0]]))
-    False
-    >>> is_sequence_of_sequences(np.array([[1, 0, 0]]))
-    False
     """
     # the explicit check for ndarray is for forward compatibility; future
     # versions of Numpy might want to register ndarray as a Sequence
@@ -213,8 +194,6 @@ def is_multilabel(y):
     >>> from sklearn.utils.multiclass import is_multilabel
     >>> is_multilabel([0, 1, 0, 1])
     False
-    >>> is_multilabel([[1], [0, 2], []])
-    True
     >>> is_multilabel(np.array([[1, 0], [0, 0]]))
     True
     >>> is_multilabel(np.array([[1], [0], [0]]))
@@ -273,10 +252,6 @@ def type_of_target(y):
     'multiclass-multioutput'
     >>> type_of_target(np.array([[1.5, 2.0], [3.0, 1.6]]))
     'continuous-multioutput'
-    >>> type_of_target([['a', 'b'], ['c'], []])
-    'multilabel-sequences'
-    >>> type_of_target([[]])
-    'multilabel-sequences'
     >>> type_of_target(np.array([[0, 1], [1, 1]]))
     'multilabel-indicator'
     """
diff --git a/sklearn/utils/tests/test_multiclass.py b/sklearn/utils/tests/test_multiclass.py
@@ -1,5 +1,6 @@
 import numpy as np
 from itertools import product
+from functools import partial
 from sklearn.externals.six.moves import xrange
 from sklearn.externals.six import iteritems
 
@@ -8,6 +9,8 @@
 from sklearn.utils.testing import assert_true
 from sklearn.utils.testing import assert_false
 from sklearn.utils.testing import assert_raises
+from sklearn.utils.testing import assert_warns
+from sklearn.utils.testing import ignore_warnings
 
 from sklearn.utils.multiclass import unique_labels
 from sklearn.utils.multiclass import is_label_indicator_matrix
@@ -132,10 +135,15 @@ def test_unique_labels():
     assert_array_equal(unique_labels([4, 0, 2]), np.array([0, 2, 4]))
 
     # Multilabels
-    assert_array_equal(unique_labels([(0, 1, 2), (0,), tuple(), (2, 1)]),
+    assert_array_equal(assert_warns(DeprecationWarning,
+                                    unique_labels,
+                                    [(0, 1, 2), (0,), tuple(), (2, 1)]),
                        np.arange(3))
-    assert_array_equal(unique_labels([[0, 1, 2], [0], list(), [2, 1]]),
+    assert_array_equal(assert_warns(DeprecationWarning,
+                                    unique_labels,
+                                    [[0, 1, 2], [0], list(), [2, 1]]),
                        np.arange(3))
+
     assert_array_equal(unique_labels(np.array([[0, 0, 1],
                                                [1, 0, 1],
                                                [0, 0, 0]])),
@@ -160,9 +168,16 @@ def test_unique_labels():
     # Some tests with strings input
     assert_array_equal(unique_labels(["a", "b", "c"], ["d"]),
                        ["a", "b", "c", "d"])
-    assert_array_equal(unique_labels([["a", "b"], ["c"]], [["d"]]),
+
+    assert_array_equal(assert_warns(DeprecationWarning, unique_labels,
+                                    [["a", "b"], ["c"]], [["d"]]),
                        ["a", "b", "c", "d"])
 
+
+@ignore_warnings
+def test_unique_labels_non_specific():
+    """Test unique_labels with a variety of collected examples"""
+
     # Smoke test for all supported format
     for format in ["binary", "multiclass", "multilabel-sequences",
                    "multilabel-indicator"]:
@@ -178,6 +193,9 @@ def test_unique_labels():
         for example in EXAMPLES[y_type]:
             assert_raises(ValueError, unique_labels, example)
 
+
+@ignore_warnings
+def test_unique_labels_mixed_types():
     #Mix of multilabel-indicator and multilabel-sequences
     mix_multilabel_format = product(EXAMPLES["multilabel-indicator"],
                                     EXAMPLES["multilabel-sequences"])
@@ -207,6 +225,7 @@ def test_unique_labels():
                        ["0", "2"])
 
 
+@ignore_warnings
 def test_is_multilabel():
     for group, group_examples in iteritems(EXAMPLES):
         if group.startswith('multilabel'):
@@ -234,14 +253,18 @@ def test_is_sequence_of_sequences():
     for group, group_examples in iteritems(EXAMPLES):
         if group == 'multilabel-sequences':
             assert_, exp = assert_true, 'True'
+            check = partial(assert_warns, DeprecationWarning,
+                            is_sequence_of_sequences)
         else:
             assert_, exp = assert_false, 'False'
+            check = is_sequence_of_sequences
         for example in group_examples:
-            assert_(is_sequence_of_sequences(example),
+            assert_(check(example),
                     msg='is_sequence_of_sequences(%r) should be %s'
                     % (example, exp))
 
 
+@ignore_warnings
 def test_type_of_target():
     for group, group_examples in iteritems(EXAMPLES):
         for example in group_examples:
