amueller
diff --git a/‎sklearn/datasets/openml.py
Lines changed: 10 additions & 1 deletion b/‎sklearn/datasets/openml.py
Lines changed: 10 additions & 1 deletion
diff --git a/‎sklearn/datasets/tests/test_openml.py
Lines changed: 7 additions & 0 deletions b/‎sklearn/datasets/tests/test_openml.py
Lines changed: 7 additions & 0 deletions
@@ -347,7 +347,7 @@ def _verify_target_data_type(features_dict, target_columns):
 
 
 def fetch_openml(name=None, version='active', data_id=None, data_home=None,
-                 target_column='default-target', cache=True):
+                 target_column='default-target', cache=True, return_X_y=False):
     """Fetch dataset from openml by name or dataset id.
 
     Datasets are uniquely identified by either an integer ID or by a
@@ -395,6 +395,10 @@ def fetch_openml(name=None, version='active', data_id=None, data_home=None,
     cache : boolean, default=True
         Whether to cache downloaded datasets using joblib.
 
+    return_X_y : boolean, default=False.
+        If True, returns ``(data, target)`` instead of a Bunch object. See
+        below for more information about the `data` and `target` objects.
+
     Returns
     -------
 
@@ -416,6 +420,8 @@ def fetch_openml(name=None, version='active', data_id=None, data_home=None,
         details : dict
             More metadata from OpenML
 
+    (data, target) : tuple if ``return_X_y`` is True
+
         .. note:: EXPERIMENTAL
 
             This interface is **experimental** as at version 0.20 and
@@ -557,6 +563,9 @@ def fetch_openml(name=None, version='active', data_id=None, data_home=None,
     elif y.shape[1] == 0:
         y = None
 
+    if return_X_y:
+        return X, y
+
     bunch = Bunch(
         data=X, target=y, feature_names=data_columns,
         DESCR=description, details=data_description,
 
@@ -15,6 +15,8 @@
                                    assert_raise_message)
 from sklearn.externals.six import string_types
 from sklearn.externals.six.moves.urllib.error import HTTPError
+from sklearn.datasets.tests.test_common import check_return_X_y
+from functools import partial
 
 
 currdir = os.path.dirname(os.path.abspath(__file__))
@@ -124,6 +126,11 @@ def _fetch_dataset_from_openml(data_id, data_name, data_version,
         # np.isnan doesn't work on CSR matrix
         assert (np.count_nonzero(np.isnan(data_by_id.data)) ==
                 expected_missing)
+
+    # test return_X_y option
+    fetch_func = partial(fetch_openml, data_id=data_id, cache=False,
+                         target_column=target_column)
+    check_return_X_y(data_by_id, fetch_func)
     return data_by_id