BUG: fix DataFrame.__getitem__ and .loc with non-list listlikes

close pandas-dev#21294
close pandas-dev#21428

Author: toobaz

doc/source/whatsnew/v0.23.1.txt

@@ -110,6 +110,8 @@ Bug Fixes
 - Bug in :meth:`Series.reset_index` where appropriate error was not raised with an invalid level name (:issue:`20925`)
 - Bug in :func:`interval_range` when ``start``/``periods`` or ``end``/``periods`` are specified with float ``start`` or ``end`` (:issue:`21161`)
 - Bug in :meth:`MultiIndex.set_names` where error raised for a ``MultiIndex`` with ``nlevels == 1`` (:issue:`21149`)
+- Bug in :meth:`DataFrame.__getitem__` and :meth:`DataFrame.loc` which did not accept columns keys passed as non-list iterables (:issue:`21294`)
+- Bug in :meth:`DataFrame.__getitem__` looking for np.nan in non-unique columns (:issue:`21428`)
 - Bug in :class:`IntervalIndex` constructors where creating an ``IntervalIndex`` from categorical data was not fully supported (:issue:`21243`, issue:`21253`)
 - Bug in :meth:`MultiIndex.sort_index` which was not guaranteed to sort correctly with ``level=1``; this was also causing data misalignment in particular :meth:`DataFrame.stack` operations (:issue:`20994`, :issue:`20945`, :issue:`21052`)
 

pandas/core/frame.py

@@ -2664,67 +2664,75 @@ def _ixs(self, i, axis=0):
     def __getitem__(self, key):
         key = com._apply_if_callable(key, self)
 
-        # shortcut if we are an actual column
-        is_mi_columns = isinstance(self.columns, MultiIndex)
+        # shortcut if the key is in columns
         try:
-            if key in self.columns and not is_mi_columns:
-                return self._getitem_column(key)
-        except:
+            if self.columns.is_unique and key in self.columns:
+                if self.columns.nlevels > 1:
+                    return self._getitem_multilevel(key)
+                return self._get_item_cache(key)
+        except (ValueError, TypeError):
             pass
 
-        # see if we can slice the rows
+        # Do we have a slicer (on rows)?
         indexer = convert_to_index_sliceable(self, key)
         if indexer is not None:
-            return self._getitem_slice(indexer)
+            return self._slice(indexer, axis=0)
 
-        if isinstance(key, (Series, np.ndarray, Index, list)):
-            # either boolean or fancy integer index
-            return self._getitem_array(key)
-        elif isinstance(key, DataFrame):
+        # Do we have a (boolean) DataFrame?
+        if isinstance(key, DataFrame):
             return self._getitem_frame(key)
-        elif is_mi_columns:
-            return self._getitem_multilevel(key)
-        else:
-            return self._getitem_column(key)
 
-    def _getitem_column(self, key):
-        """ return the actual column """
+        # Do we have a (boolean) 1d indexer?
+        if com.is_bool_indexer(key):
+            return self._getitem_bool_array(key)
+
+        # We are left with two options: a single key, and a collection of keys,
+        # We interpret tuples as collections only for non-MultiIndex
+        is_single_key = isinstance(key, tuple) or not is_list_like(key)
+
+        if is_single_key:
+            if self.columns.nlevels > 1:
+                return self._getitem_multilevel(key)
+            indexer = self.columns.get_loc(key)
+            if is_integer(indexer):
+                indexer = [indexer]
+        else:
+            indexer = self.loc._convert_to_indexer(key, axis=1)
 
-        # get column
-        if self.columns.is_unique:
-            return self._get_item_cache(key)
+        # take() does not accept boolean indexers
+        if getattr(indexer, "dtype", None) == bool:
+            indexer = np.where(indexer)[0]
 
-        # duplicate columns & possible reduce dimensionality
-        result = self._constructor(self._data.get(key))
-        if result.columns.is_unique:
-            result = result[key]
+        data = self._take(indexer, axis=1)
 
-        return result
+        if is_single_key:
+            # What does looking for a single key in a non-unique index return?
+            # The behavior is inconsistent. It returns a Series, except when
+            # - the key itself is repeated (test on data.shape, #9519), or
+            # - we have a MultiIndex on columns (test on self.columns, #21309)
+            if data.shape[1] == 1 and not isinstance(self.columns, MultiIndex):
+                data = data[key]
 
-    def _getitem_slice(self, key):
-        return self._slice(key, axis=0)
+        return data
 
-    def _getitem_array(self, key):
+    def _getitem_bool_array(self, key):
         # also raises Exception if object array with NA values
-        if com.is_bool_indexer(key):
-            # warning here just in case -- previously __setitem__ was
-            # reindexing but __getitem__ was not; it seems more reasonable to
-            # go with the __setitem__ behavior since that is more consistent
-            # with all other indexing behavior
-            if isinstance(key, Series) and not key.index.equals(self.index):
-                warnings.warn("Boolean Series key will be reindexed to match "
-                              "DataFrame index.", UserWarning, stacklevel=3)
-            elif len(key) != len(self.index):
-                raise ValueError('Item wrong length %d instead of %d.' %
-                                 (len(key), len(self.index)))
-            # check_bool_indexer will throw exception if Series key cannot
-            # be reindexed to match DataFrame rows
-            key = check_bool_indexer(self.index, key)
-            indexer = key.nonzero()[0]
-            return self._take(indexer, axis=0)
-        else:
-            indexer = self.loc._convert_to_indexer(key, axis=1)
-            return self._take(indexer, axis=1)
+        # warning here just in case -- previously __setitem__ was
+        # reindexing but __getitem__ was not; it seems more reasonable to
+        # go with the __setitem__ behavior since that is more consistent
+        # with all other indexing behavior
+        if isinstance(key, Series) and not key.index.equals(self.index):
+            warnings.warn("Boolean Series key will be reindexed to match "
+                          "DataFrame index.", UserWarning, stacklevel=3)
+        elif len(key) != len(self.index):
+            raise ValueError('Item wrong length %d instead of %d.' %
+                             (len(key), len(self.index)))
+
+        # check_bool_indexer will throw exception if Series key cannot
+        # be reindexed to match DataFrame rows
+        key = check_bool_indexer(self.index, key)
+        indexer = key.nonzero()[0]
+        return self._take(indexer, axis=0)
 
     def _getitem_multilevel(self, key):
         loc = self.columns.get_loc(key)

pandas/tests/frame/test_indexing.py

@@ -92,45 +92,46 @@ def test_get(self):
             result = df.get(None)
             assert result is None
 
-    def test_getitem_iterator(self):
+    def test_loc_iterable(self):
         idx = iter(['A', 'B', 'C'])
         result = self.frame.loc[:, idx]
         expected = self.frame.loc[:, ['A', 'B', 'C']]
         assert_frame_equal(result, expected)
 
-        idx = iter(['A', 'B', 'C'])
-        result = self.frame.loc[:, idx]
-        expected = self.frame.loc[:, ['A', 'B', 'C']]
-        assert_frame_equal(result, expected)
+    @pytest.mark.parametrize(
+        "idx_type",
+        [list, iter, Index, set,
+         lambda l: dict(zip(l, range(len(l)))),
+         lambda l: dict(zip(l, range(len(l)))).keys()],
+        ids=["list", "iter", "Index", "set", "dict", "dict_keys"])
+    @pytest.mark.parametrize("levels", [1, 2])
+    def test_getitem_listlike(self, idx_type, levels):
+        # GH 21294
+
+        if levels == 1:
+            frame, missing = self.frame, 'food'
+        else:
+            # MultiIndex columns
+            frame = DataFrame(randn(8, 3),
+                              columns=Index([('foo', 'bar'), ('baz', 'qux'),
+                                             ('peek', 'aboo')],
+                                            name=('sth', 'sth2')))
+            missing = ('good', 'food')
 
-    def test_getitem_list(self):
-        self.frame.columns.name = 'foo'
+        keys = [frame.columns[1], frame.columns[0]]
+        idx = idx_type(keys)
+        idx_check = list(idx_type(keys))
 
-        result = self.frame[['B', 'A']]
-        result2 = self.frame[Index(['B', 'A'])]
+        result = frame[idx]
 
-        expected = self.frame.loc[:, ['B', 'A']]
-        expected.columns.name = 'foo'
+        expected = frame.loc[:, idx_check]
+        expected.columns.names = frame.columns.names
 
         assert_frame_equal(result, expected)
-        assert_frame_equal(result2, expected)
 
-        assert result.columns.name == 'foo'
-
-        with tm.assert_raises_regex(KeyError, 'not in index'):
-            self.frame[['B', 'A', 'food']]
+        idx = idx_type(keys + [missing])
         with tm.assert_raises_regex(KeyError, 'not in index'):
-            self.frame[Index(['B', 'A', 'foo'])]
-
-        # tuples
-        df = DataFrame(randn(8, 3),
-                       columns=Index([('foo', 'bar'), ('baz', 'qux'),
-                                      ('peek', 'aboo')], name=('sth', 'sth2')))
-
-        result = df[[('foo', 'bar'), ('baz', 'qux')]]
-        expected = df.iloc[:, :2]
-        assert_frame_equal(result, expected)
-        assert result.columns.names == ('sth', 'sth2')
+            frame[idx]
 
     def test_getitem_callable(self):
         # GH 12533