FIX remove duplicates in MultiLabelBinarizer

jnothman · jnothman · commit 2ab7c62df3bd · 2013-12-12T09:03:54.000+11:00
diff --git a/sklearn/preprocessing/label.py b/sklearn/preprocessing/label.py
@@ -558,7 +558,7 @@ def _transform(self, y, class_mapping):
         indices = array.array('i')
         indptr = array.array('i', [0])
         for labels in y:
-            indices.extend(class_mapping[label] for label in labels)
+            indices.extend(set(class_mapping[label] for label in labels))
             indptr.append(len(indices))
         data = np.ones(len(indices), dtype=int)
         return sp.csr_matrix((data, indices, indptr),
diff --git a/sklearn/preprocessing/tests/test_label.py b/sklearn/preprocessing/tests/test_label.py
@@ -371,3 +371,12 @@ def test_mutlilabel_binarizer_non_integer_labels():
 
     lsb = MultiLabelBinarizer()
     assert_raises(TypeError, lsb.fit_transform, [({}), ({}, {'a': 'b'})])
+
+
+def test_mutlilabel_binarizer_non_unique():
+    inp = [(1, 1, 1, 0)]
+    indicator_mat = np.array([[1, 1]])
+    lsb = MultiLabelBinarizer()
+    assert_array_equal(lsb.fit_transform(inp), indicator_mat)
+
+    assert_array_equal(lsb.inverse_transform(np.array([[1, 3]])), [(0, 1,)])
