BUG: fix KeyError with list of a single, missing, element (#27154)

closes #27148

Author: toobaz

doc/source/whatsnew/v0.25.0.rst

@@ -1046,6 +1046,7 @@ Indexing
 - Bug in which :meth:`DataFrame.to_csv` caused a segfault for a reindexed data frame, when the indices were single-level :class:`MultiIndex` (:issue:`26303`).
 - Fixed bug where assigning a :class:`arrays.PandasArray` to a :class:`pandas.core.frame.DataFrame` would raise error (:issue:`26390`)
 - Allow keyword arguments for callable local reference used in the :meth:`DataFrame.query` string (:issue:`26426`)
+- Fixed a ``KeyError`` when indexing a :class:`MultiIndex`` level with a list containing exactly one label, which is missing (:issue:`27148`)
 - Bug which produced ``AttributeError`` on partial matching :class:`Timestamp` in a :class:`MultiIndex`  (:issue:`26944`)
 - Bug in :class:`Categorical` and  :class:`CategoricalIndex` with :class:`Interval` values when using the ``in`` operator (``__contains``) with objects that are not comparable to the values in the ``Interval`` (:issue:`23705`)
 - Bug in :meth:`DataFrame.loc` and :meth:`DataFrame.iloc` on a :class:`DataFrame` with a single timezone-aware datetime64[ns] column incorrectly returning a scalar instead of a :class:`Series` (:issue:`27110`)

pandas/core/indexing.py

@@ -1861,7 +1861,7 @@ def _getitem_axis(self, key, axis=None):
 
                 if (
                     not isinstance(key, tuple)
-                    and len(key) > 1
+                    and len(key)
                     and not isinstance(key[0], tuple)
                 ):
                     key = tuple([key])

pandas/tests/indexing/multiindex/test_loc.py

@@ -130,6 +130,19 @@ def test_loc_multiindex_missing_label_raises(self):
         with pytest.raises(KeyError, match=r"^2$"):
             df.loc[2]
 
+    @pytest.mark.parametrize("key, pos", [([2, 4], [0, 1]), ([2], []), ([2, 3], [])])
+    def test_loc_multiindex_list_missing_label(self, key, pos):
+        # GH 27148 - lists with missing labels do not raise:
+        df = DataFrame(
+            np.random.randn(3, 3),
+            columns=[[2, 2, 4], [6, 8, 10]],
+            index=[[4, 4, 8], [8, 10, 12]],
+        )
+
+        expected = df.iloc[pos]
+        result = df.loc[key]
+        tm.assert_frame_equal(result, expected)
+
     def test_loc_multiindex_too_many_dims_raises(self):
         # GH 14885
         s = Series(
@@ -280,47 +293,27 @@ def convert_nested_indexer(indexer_type, keys):
 
 
 @pytest.mark.parametrize(
-    "indexer, is_level1, expected_error",
+    "indexer, pos",
     [
-        ([], False, None),  # empty ok
-        (["A"], False, None),
-        (["A", "D"], False, None),
-        (["D"], False, r"\['D'\] not in index"),  # not any values found
-        (pd.IndexSlice[:, ["foo"]], True, None),
-        (pd.IndexSlice[:, ["foo", "bah"]], True, None),
+        ([], []),  # empty ok
+        (["A"], slice(3)),
+        (["A", "D"], slice(3)),
+        (["D", "E"], []),  # no values found - fine
+        (["D"], []),  # same, with single item list: GH 27148
+        (pd.IndexSlice[:, ["foo"]], slice(2, None, 3)),
+        (pd.IndexSlice[:, ["foo", "bah"]], slice(2, None, 3)),
     ],
 )
-def test_loc_getitem_duplicates_multiindex_missing_indexers(
-    indexer, is_level1, expected_error
-):
+def test_loc_getitem_duplicates_multiindex_missing_indexers(indexer, pos):
     # GH 7866
     # multi-index slicing with missing indexers
     idx = MultiIndex.from_product(
         [["A", "B", "C"], ["foo", "bar", "baz"]], names=["one", "two"]
     )
     s = Series(np.arange(9, dtype="int64"), index=idx).sort_index()
-
-    if indexer == []:
-        expected = s.iloc[[]]
-    elif is_level1:
-        expected = Series(
-            [0, 3, 6],
-            index=MultiIndex.from_product(
-                [["A", "B", "C"], ["foo"]], names=["one", "two"]
-            ),
-        ).sort_index()
-    else:
-        exp_idx = MultiIndex.from_product(
-            [["A"], ["foo", "bar", "baz"]], names=["one", "two"]
-        )
-        expected = Series(np.arange(3, dtype="int64"), index=exp_idx).sort_index()
-
-    if expected_error is not None:
-        with pytest.raises(KeyError, match=expected_error):
-            s.loc[indexer]
-    else:
-        result = s.loc[indexer]
-        tm.assert_series_equal(result, expected)
+    expected = s.iloc[pos]
+    result = s.loc[indexer]
+    tm.assert_series_equal(result, expected)
 
 
 def test_series_loc_getitem_fancy(multiindex_year_month_day_dataframe_random_data):

pandas/tests/indexing/multiindex/test_slice.py

@@ -117,11 +117,11 @@ def test_per_axis_per_level_getitem(self):
         with pytest.raises(ValueError):
             df.loc[(slice(None), np.array([True, False])), :]
 
-        # ambiguous cases
-        # these can be multiply interpreted (e.g. in this case
-        # as df.loc[slice(None),[1]] as well
-        with pytest.raises(KeyError, match=r"'\[1\] not in index'"):
-            df.loc[slice(None), [1]]
+        # ambiguous notation
+        # this is interpreted as slicing on both axes (GH #16396)
+        result = df.loc[slice(None), [1]]
+        expected = df.iloc[:, []]
+        tm.assert_frame_equal(result, expected)
 
         result = df.loc[(slice(None), [1]), :]
         expected = df.iloc[[0, 3]]