BUG: Fix groupby over a CategoricalIndex in axis=1

closes GH18432

Author: Eric Kisslinger

pandas/core/groupby.py

@@ -2859,7 +2859,7 @@ def is_in_obj(gpr):
         else:
             in_axis, name = False, None
 
-        if is_categorical_dtype(gpr) and len(gpr) != len(obj):
+        if is_categorical_dtype(gpr) and len(gpr) != obj.shape[axis]:
             raise ValueError("Categorical dtype grouper must "
                              "have len(grouper) == len(data)")
 

pandas/tests/groupby/test_categorical.py

@@ -191,6 +191,25 @@ def test_groupby_categorical_index(self):
                 [0, 1, 2, 3], levels, ordered=True), name='cats')
         assert_frame_equal(result, expected)
 
+    def test_groupby_categorical_columns_index(self):
+        # GH18432
+        columns = ['A', 'B', 'A', 'B']
+        categories = ['B', 'A']
+        data = np.random.rand(10, 4)
+        cat_index = CategoricalIndex(columns,
+                                     categories=categories,
+                                     ordered=True)
+        df = DataFrame(data, columns=cat_index)
+        result = df.groupby(axis=1, level=0).sum()
+        df = DataFrame(data, columns=columns)
+        expected = df.groupby(axis=1, level=0).sum()
+        expected = expected[categories]
+        expected.columns = CategoricalIndex(categories,
+                                            categories=categories,
+                                            ordered=True)
+
+        assert_frame_equal(result, expected)
+
     def test_groupby_describe_categorical_columns(self):
         # GH 11558
         cats = pd.CategoricalIndex(['qux', 'foo', 'baz', 'bar'],