pandas-dev
diff --git a/‎doc/source/release.rst‎
Lines changed: 4 additions & 0 deletions b/‎doc/source/release.rst‎
Lines changed: 4 additions & 0 deletions
diff --git a/‎pandas/core/frame.py‎
Lines changed: 6 additions & 1 deletion b/‎pandas/core/frame.py‎
Lines changed: 6 additions & 1 deletion
diff --git a/‎pandas/tests/test_frame.py‎
Lines changed: 16 additions & 3 deletions b/‎pandas/tests/test_frame.py‎
Lines changed: 16 additions & 3 deletions
@@ -424,6 +424,10 @@ Bug Fixes
     across different versions of matplotlib (:issue:`4789`)
   - Suppressed DeprecationWarning associated with internal calls issued by repr() (:issue:`4391`)
   - Fixed an issue with a duplicate index and duplicate selector with ``.loc`` (:issue:`4825`)
+  - Fixed an issue with ``DataFrame.sort_index`` where, when sorting by a
+    single column and passing a list for ``ascending``, the argument for
+    ``ascending`` was being interpreted as ``True`` (:issue:`4839`,
+    :issue:`4846`)
 
 pandas 0.12.0
 -------------
 
@@ -2856,7 +2856,7 @@ def sort_index(self, axis=0, by=None, ascending=True, inplace=False,
 
         Examples
         --------
-        >>> result = df.sort_index(by=['A', 'B'], ascending=[1, 0])
+        >>> result = df.sort_index(by=['A', 'B'], ascending=[True, False])
 
         Returns
         -------
@@ -2875,6 +2875,9 @@ def sort_index(self, axis=0, by=None, ascending=True, inplace=False,
                 raise ValueError('When sorting by column, axis must be 0 (rows)')
             if not isinstance(by, (tuple, list)):
                 by = [by]
+            if com._is_sequence(ascending) and len(by) != len(ascending):
+                raise ValueError('Length of ascending (%d) != length of by'
+                                 ' (%d)' % (len(ascending), len(by)))
 
             if len(by) > 1:
                 keys = []
@@ -2900,6 +2903,8 @@ def trans(v):
                     raise ValueError('Cannot sort by duplicate column %s'
                                      % str(by))
                 indexer = k.argsort(kind=kind)
+                if isinstance(ascending, (tuple, list)):
+                    ascending = ascending[0]
                 if not ascending:
                     indexer = indexer[::-1]
         elif isinstance(labels, MultiIndex):
 
@@ -8796,24 +8796,37 @@ def test_sort_index(self):
         expected = frame.ix[frame.index[indexer]]
         assert_frame_equal(sorted_df, expected)
 
+        sorted_df = frame.sort(columns='A', ascending=False)
+        assert_frame_equal(sorted_df, expected)
+
+        # GH4839
+        sorted_df = frame.sort(columns=['A'], ascending=[False])
+        assert_frame_equal(sorted_df, expected)
+
         # check for now
         sorted_df = frame.sort(columns='A')
+        assert_frame_equal(sorted_df, expected[::-1])
         expected = frame.sort_index(by='A')
         assert_frame_equal(sorted_df, expected)
 
-        sorted_df = frame.sort(columns='A', ascending=False)
-        expected = frame.sort_index(by='A', ascending=False)
-        assert_frame_equal(sorted_df, expected)
 
         sorted_df = frame.sort(columns=['A', 'B'], ascending=False)
         expected = frame.sort_index(by=['A', 'B'], ascending=False)
         assert_frame_equal(sorted_df, expected)
 
+        sorted_df = frame.sort(columns=['A', 'B'])
+        assert_frame_equal(sorted_df, expected[::-1])
+
         self.assertRaises(ValueError, frame.sort_index, axis=2, inplace=True)
+
         msg = 'When sorting by column, axis must be 0'
         with assertRaisesRegexp(ValueError, msg):
             frame.sort_index(by='A', axis=1)
 
+        msg = r'Length of ascending \(5\) != length of by \(2\)'
+        with assertRaisesRegexp(ValueError, msg):
+            frame.sort_index(by=['A', 'B'], axis=0, ascending=[True] * 5)
+
     def test_sort_index_multicolumn(self):
         import random
         A = np.arange(5).repeat(20)