Merge pull request #7107 from jreback/mi_slicers

BUG: remove unique requirement rom MultiIndex.get_locs (GH7106)

Author: jreback

doc/source/release.rst

@@ -295,7 +295,7 @@ Improvements to existing features
   the func (:issue:`6289`)
 - ``plot(legend='reverse')`` will now reverse the order of legend labels for most plot kinds.
   (:issue:`6014`)
-- Allow multi-index slicers (:issue:`6134`, :issue:`4036`, :issue:`3057`, :issue:`2598`, :issue:`5641`)
+- Allow multi-index slicers (:issue:`6134`, :issue:`4036`, :issue:`3057`, :issue:`2598`, :issue:`5641`, :issue:`7106`)
 - improve performance of slice indexing on Series with string keys (:issue:`6341`, :issue:`6372`)
 - implement joining a single-level indexed DataFrame on a matching column of a multi-indexed DataFrame (:issue:`3662`)
 - Performance improvement in indexing into a multi-indexed Series (:issue:`5567`)

doc/source/v0.14.0.txt

@@ -290,7 +290,7 @@ You can use ``slice(None)`` to select all the contents of *that* level. You do n
 As usual, **both sides** of the slicers are included as this is label indexing.
 
 See :ref:`the docs<indexing.mi_slicers>`
-See also issues (:issue:`6134`, :issue:`4036`, :issue:`3057`, :issue:`2598`, :issue:`5641`)
+See also issues (:issue:`6134`, :issue:`4036`, :issue:`3057`, :issue:`2598`, :issue:`5641`, :issue:`7106`)
 
 .. warning::
 

pandas/core/index.py

@@ -2595,7 +2595,7 @@ def get_level_values(self, level):
         unique = self.levels[num]  # .values
         labels = self.labels[num]
         filled = com.take_1d(unique.values, labels, fill_value=unique._na_value)
-        values = unique._simple_new(filled, self.names[num], 
+        values = unique._simple_new(filled, self.names[num],
                                     freq=getattr(unique, 'freq', None),
                                     tz=getattr(unique, 'tz', None))
         return values
@@ -3556,8 +3556,6 @@ def get_locs(self, tup):
         if not self.is_lexsorted_for_tuple(tup):
             raise KeyError('MultiIndex Slicing requires the index to be fully lexsorted'
                            ' tuple len ({0}), lexsort depth ({1})'.format(len(tup), self.lexsort_depth))
-        if not self.is_unique:
-            raise ValueError('MultiIndex Slicing requires a unique index')
 
         def _convert_indexer(r):
             if isinstance(r, slice):

pandas/tests/test_indexing.py

@@ -1356,6 +1356,47 @@ def f():
             df.loc[(slice(None),[1])]
         self.assertRaises(KeyError, f)
 
+        # not lexsorted
+        self.assertEquals(df.index.lexsort_depth,2)
+        df = df.sortlevel(level=1,axis=0)
+        self.assertEquals(df.index.lexsort_depth,0)
+        with tm.assertRaisesRegexp(KeyError, 'MultiIndex Slicing requires the index to be fully lexsorted tuple len \(2\), lexsort depth \(0\)'):
+            df.loc[(slice(None),df.loc[:,('a','bar')]>5),:]
+
+    def test_multiindex_slicers_non_unique(self):
+
+        # GH 7106
+        # non-unique mi index support
+        df = DataFrame(dict(A = ['foo','foo','foo','foo'],
+                            B = ['a','a','a','a'],
+                            C = [1,2,1,3],
+                            D = [1,2,3,4])).set_index(['A','B','C']).sortlevel()
+        self.assertFalse(df.index.is_unique)
+        expected = DataFrame(dict(A = ['foo','foo'],
+                                  B = ['a','a'],
+                                  C = [1,1],
+                                  D = [1,3])).set_index(['A','B','C']).sortlevel()
+        result = df.loc[(slice(None),slice(None),1),:]
+        assert_frame_equal(result, expected)
+
+        # this is equivalent of an xs expression
+        result = df.xs(1,level=2,drop_level=False)
+        assert_frame_equal(result, expected)
+
+        df = DataFrame(dict(A = ['foo','foo','foo','foo'],
+                            B = ['a','a','a','a'],
+                            C = [1,2,1,2],
+                            D = [1,2,3,4])).set_index(['A','B','C']).sortlevel()
+        self.assertFalse(df.index.is_unique)
+        expected = DataFrame(dict(A = ['foo','foo'],
+                                  B = ['a','a'],
+                                  C = [1,1],
+                                  D = [1,3])).set_index(['A','B','C']).sortlevel()
+        result = df.loc[(slice(None),slice(None),1),:]
+        self.assertFalse(result.index.is_unique)
+        assert_frame_equal(result, expected)
+
+
     def test_per_axis_per_level_doc_examples(self):
 
         # test index maker