BUG: don't attach nonsense 'result' name to groupby results without obvious name, close #995

Author: wesm

RELEASE.rst

@@ -32,6 +32,7 @@ pandas 0.7.3
   - Add group_keys argument to groupby to not add group names to MultiIndex in
     result of apply (GH #938)
   - scatter_matrix method in pandas/tools/plotting.py (PR #935)
+  - Add ``kurt`` methods to Series and DataFrame (PR #964)
 
 **API Changes**
 
@@ -40,6 +41,8 @@ pandas 0.7.3
 
 **Bug fixes**
 
+  - Don't attach nonsense 'result' name to groupby results (GH #995)
+
 pandas 0.7.2
 ============
 

pandas/core/groupby.py

@@ -131,7 +131,7 @@ def indices(self):
     @property
     def name(self):
         if self._column is None:
-            return 'result'
+            return None # 'result'
         else:
             return self._column
 

pandas/tests/test_groupby.py

@@ -1606,6 +1606,14 @@ def test_skip_group_keys(self):
         expected = concat(pieces)
         assert_series_equal(result, expected)
 
+    def test_no_nonsense_name(self):
+        # GH #995
+        s = self.frame['C'].copy()
+        s.name = None
+
+        result = s.groupby(self.frame['A']).agg(np.sum)
+        self.assert_(result.name is None)
+
 def _check_groupby(df, result, keys, field, f=lambda x: x.sum()):
     tups = map(tuple, df[keys].values)
     tups = com._asarray_tuplesafe(tups)