BUG: Bug upon Series.Groupby.nunique with empty Series

closes #12553
closes #14770

(cherry picked from commit c0e13d1)

Author: mroeschke

doc/source/whatsnew/v0.19.2.txt

@@ -59,6 +59,8 @@ Bug Fixes
 
 
 - Bug ``HDFStore`` writing a ``MultiIndex`` when using ``data_columns=True`` (:issue:`14435`)
+- Bug in ``Series.groupby.nunique()`` raising an ``IndexError`` for an empty ``Series`` (:issue:`12553`)
+
 
 
 - Bug in clipboard functions on linux with python2 with unicode and separators (:issue:`13747`)

pandas/core/groupby.py

@@ -2908,6 +2908,7 @@ def true_and_notnull(x, *args, **kwargs):
     def nunique(self, dropna=True):
         """ Returns number of unique elements in the group """
         ids, _, _ = self.grouper.group_info
+
         val = self.obj.get_values()
 
         try:
@@ -2938,7 +2939,10 @@ def nunique(self, dropna=True):
             inc[idx] = 1
 
         out = np.add.reduceat(inc, idx).astype('int64', copy=False)
-        res = out if ids[0] != -1 else out[1:]
+        if len(ids):
+            res = out if ids[0] != -1 else out[1:]
+        else:
+            res = out[1:]
         ri = self.grouper.result_index
 
         # we might have duplications among the bins

pandas/tests/test_groupby.py

@@ -6773,6 +6773,13 @@ def test_nunique_with_object(self):
         expected = pd.Series([1] * 5, name='name', index=index)
         tm.assert_series_equal(result, expected)
 
+    def test_nunique_with_empty_series(self):
+        # GH 12553
+        data = pd.Series(name='name')
+        result = data.groupby(level=0).nunique()
+        expected = pd.Series(name='name', dtype='int64')
+        tm.assert_series_equal(result, expected)
+
     def test_transform_with_non_scalar_group(self):
         # GH 10165
         cols = pd.MultiIndex.from_tuples([