TST label_binarize_dense_output

sklearn/preprocessing/label.py

@@ -6,7 +6,7 @@
 
 import numpy as np
 
-from scipy.sparse import coo_matrix
+from scipy.sparse import coo_matrix, issparse
 
 from ..base import BaseEstimator, TransformerMixin
 
@@ -170,6 +170,11 @@ class LabelBinarizer(BaseEstimator, TransformerMixin):
     pos_label : int (default: 1)
         Value with which positive labels must be encoded.
 
+    dense_output : boolean (default: False)
+        If True, ensure that the output of label_binarize is a
+        dense numpy array even if the binarize matrix is sparse.
+        If False, the binarized data use a sparse representation.
+
     Attributes
     ----------
     `classes_` : array of shape [n_class]
@@ -207,12 +212,13 @@ class LabelBinarizer(BaseEstimator, TransformerMixin):
         LabelBinarizer with fixed classes.
     """
 
-    def __init__(self, neg_label=0, pos_label=1):
+    def __init__(self, neg_label=0, pos_label=1, dense_output=True):
         if neg_label >= pos_label:
             raise ValueError("neg_label must be strictly less than pos_label.")
 
         self.neg_label = neg_label
         self.pos_label = pos_label
+        self.dense_output = dense_output
 
     @property
     @deprecated("Attribute `multilabel` was renamed to `multilabel_` in "
@@ -260,7 +266,7 @@ def transform(self, y):
 
         Returns
         -------
-        Y : numpy array of shape [n_samples, n_classes]
+        Y : numpy array or COO matrix of shape [n_samples, n_classes]
         """
         self._check_fitted()
 
@@ -271,9 +277,9 @@ def transform(self, y):
                              " input.")
 
         return label_binarize(y, self.classes_,
-                              multilabel=self.multilabel_,
                               pos_label=self.pos_label,
-                              neg_label=self.neg_label)
+                              neg_label=self.neg_label,
+                              dense_output=self.dense_output)
 
     def inverse_transform(self, Y, threshold=None):
         """Transform binary labels back to multi-class labels
@@ -315,13 +321,13 @@ def inverse_transform(self, Y, threshold=None):
             threshold = self.neg_label + half
 
         if self.multilabel_:
-            if not(isinstance(Y, coo_matrix)):
+            if not(issparse(Y)):
                 Y = np.array(Y > threshold, dtype=int)
             # Return the predictions in the same format as in fit
             if self.indicator_matrix_:
                 # Label indicator matrix format
                 return Y
-            elif isinstance(Y, coo_matrix):
+            elif issparse(Y):
                 # Splitting and processing sparse matrix
                 y = []
                 for i in range(Y.shape[0]):
@@ -404,7 +410,7 @@ def label_binarize(y, classes, neg_label=0, pos_label=1,
         raise ValueError("neg_label cannot equal pos_label")
 
     y_type = type_of_target(y)
-    n_samples = len(y)
+    n_samples = y.shape[0] if issparse(y) else len(y)
     n_classes = len(classes)
     classes = np.asarray(classes)
     sorted_class = np.sort(classes)

sklearn/preprocessing/tests/test_label.py

@@ -11,6 +11,7 @@
 from sklearn.preprocessing.label import LabelEncoder
 from sklearn.preprocessing.label import label_binarize
 
+from scipy.sparse import coo_matrix, csc_matrix, csr_matrix
 
 from sklearn import datasets
 from sklearn.linear_model.stochastic_gradient import SGDClassifier
@@ -147,6 +148,17 @@ def test_label_binarizer_multilabel():
     assert_equal([set(x) for x in lb.inverse_transform(got)],
                  [set(x) for x in inp])
 
+def test_label_binarizer_sparse_rep():
+        # TODO !!!
+
+
+        lb = LabelBinarizer(neg_label=neg_label, pos_label=pos_label,
+                            dense_output=dense_output)
+        output = lb.fit_transform(y)
+        assert_array_equal(toarray(output), expected)
+        inverse_output = lb.inverse_transform(output)
+        assert_array_equal(toarray(inverse_output), y)
+
 
 def test_label_binarizer_errors():
     """Check that invalid arguments yield ValueError"""
@@ -164,6 +176,8 @@ def test_label_binarizer_errors():
     assert_raises(ValueError, LabelBinarizer, neg_label=2, pos_label=1)
     assert_raises(ValueError, LabelBinarizer, neg_label=2, pos_label=2)
 
+    lb = LabelBinarizer(neg_label=-1, dense_output=False)
+    assert_raises(ValueError, lb.transform, [(1,0,2)])
 
 def test_label_encoder():
     """Test LabelEncoder's transform and inverse_transform methods"""
@@ -253,3 +267,66 @@ def test_label_binarize_with_multilabel_indicator():
 
     output = lb.fit(y).transform(y)
     assert_array_equal(output, expected)
+
+############
+
+def test_label_binarize_with_class_order():
+    out = label_binarize([1, 6], classes=[1, 2, 4, 6])
+    expected = np.array([[1, 0, 0, 0], [0, 0, 0, 1]])
+    assert_array_equal(out, expected)
+
+    # Modified class order
+    out = label_binarize([1, 6], classes=[1, 6, 4, 2])
+    expected = np.array([[1, 0, 0, 0], [0, 1, 0, 0]])
+    assert_array_equal(out, expected)
+
+def check_dense_output(y, classes, pos_label, neg_label, expected):
+    for dense_output in [True, False]:
+        output = label_binarize(y, classes, neg_label=neg_label,
+                                pos_label=pos_label, dense_output=dense_output)
+        assert_array_equal(toarray(output), expected)
+
+def test_label_binarize_dense_output_binary():
+    y = [0, 1, 0]
+    classes = [0, 1]
+    pos_label = 2
+    neg_label = -1
+    expected = np.array([[2, -1], [-1, 2], [2, -1]])[:, 1].reshape((-1, 1))
+
+    yield check_dense_output, y, classes, pos_label, neg_label, expected
+
+
+def test_label_binarize_dense_output_multiclass():
+    y = [0, 1, 2]
+    classes = [0, 1, 2]
+    pos_label = 2
+    neg_label = 0
+    expected = 2 * np.eye(3)
+
+    yield check_dense_output, y, classes, pos_label, neg_label, expected
+
+    assert_raises(ValueError, label_binarize, y, classes, neg_label=-1,
+                  pos_label=pos_label, dense_output=False)
+
+
+def test_label_binarize_dense_output_multilabel():
+    y_seq = [(1,), (0, 1, 2), tuple()]
+    y_seq_repeated = [(1, 1,), (1, 0, 2, 2), tuple()]
+    y_ind = np.array([[0, 1, 0], [1, 1, 1], [0, 0, 0]])
+    classes = [0, 1, 2]
+    pos_label = 2
+    neg_label = 0
+    expected = pos_label * y_ind
+    case_list = [y_seq, y_ind, y_seq_repeated, coo_matrix(y_ind),
+                 csr_matrix(y_ind), csc_matrix(y_ind)]
+    for y in case_list:
+        yield check_dense_output, y, classes, pos_label, neg_label, expected
+
+    assert_raises(ValueError, label_binarize, y, classes, neg_label=-1,
+                  pos_label=pos_label, dense_output=False)
+
+def test_label_binarize_errors():
+    y = [1,2,3,1]
+    classes = [0,1,2,3]
+    assert_raises(ValueError, label_binarize, y, classes, neg_label=1,
+                  pos_label=1)