Merge pull request #8417 from jreback/mi_grouping

BUG: regression in groupby with a pass thru multiindex on axis=1 (GH7997)

Author: jreback

doc/source/v0.15.0.txt

@@ -549,7 +549,7 @@ Internal Refactoring
 
 In 0.15.0 ``Index`` has internally been refactored to no longer sub-class ``ndarray``
 but instead subclass ``PandasObject``, similarly to the rest of the pandas objects. This change allows very easy sub-classing and creation of new index types. This should be
-a transparent change with only very limited API implications (:issue:`5080`, :issue:`7439`, :issue:`7796`, :issue:`8024`, :issue:`8367`)
+a transparent change with only very limited API implications (:issue:`5080`, :issue:`7439`, :issue:`7796`, :issue:`8024`, :issue:`8367`, :issue:`7997`)
 
 - you may need to unpickle pandas version < 0.15.0 pickles using ``pd.read_pickle`` rather than ``pickle.load``. See :ref:`pickle docs <io.pickle>`
 - when plotting with a ``PeriodIndex``. The ``matplotlib`` internal axes will now be arrays of ``Period`` rather than a ``PeriodIndex``. (this is similar to how a ``DatetimeIndex`` passes arrays of ``datetimes`` now)

pandas/core/groupby.py

@@ -3233,7 +3233,8 @@ def _reindex_output(self, result):
 
         levels_list = [ ping._group_index for ping in groupings ]
         index = MultiIndex.from_product(levels_list, names=self.grouper.names)
-        return result.reindex(**{ self.obj._get_axis_name(self.axis) : index, 'copy' : False }).sortlevel()
+        d = { self.obj._get_axis_name(self.axis) : index, 'copy' : False }
+        return result.reindex(**d).sortlevel(axis=self.axis)
 
     def _iterate_column_groupbys(self):
         for i, colname in enumerate(self._selected_obj.columns):

pandas/tests/test_groupby.py

@@ -1495,6 +1495,16 @@ def test_groupby_as_index_agg(self):
         result3 = grouped['C'].agg({'Q': np.sum})
         assert_frame_equal(result3, expected3)
 
+    def test_mulitindex_passthru(self):
+
+        # GH 7997
+        # regression from 0.14.1
+        df = pd.DataFrame([[1,2,3],[4,5,6],[7,8,9]])
+        df.columns = pd.MultiIndex.from_tuples([(0,1),(1,1),(2,1)])
+
+        result = df.groupby(axis=1, level=[0,1]).first()
+        assert_frame_equal(result, df)
+
     def test_multifunc_select_col_integer_cols(self):
         df = self.df
         df.columns = np.arange(len(df.columns))