REGR: Series access with Index of tuples/frozenset (#36147)

Author: rhshadrach

doc/source/whatsnew/v1.1.3.rst

@@ -14,6 +14,8 @@ including other versions of pandas.
 
 Fixed regressions
 ~~~~~~~~~~~~~~~~~
+- Fixed regression in :meth:`Series.__getitem__` incorrectly raising when the input was a tuple (:issue:`35534`)
+- Fixed regression in :meth:`Series.__getitem__` incorrectly raising when the input was a frozenset (:issue:`35747`)
 -
 
 .. ---------------------------------------------------------------------------

pandas/core/series.py

@@ -889,21 +889,19 @@ def __getitem__(self, key):
         elif key_is_scalar:
             return self._get_value(key)
 
-        if (
-            isinstance(key, tuple)
-            and is_hashable(key)
-            and isinstance(self.index, MultiIndex)
-        ):
+        if is_hashable(key):
             # Otherwise index.get_value will raise InvalidIndexError
             try:
+                # For labels that don't resolve as scalars like tuples and frozensets
                 result = self._get_value(key)
 
                 return result
 
             except KeyError:
-                # We still have the corner case where this tuple is a key
-                #  in the first level of our MultiIndex
-                return self._get_values_tuple(key)
+                if isinstance(key, tuple) and isinstance(self.index, MultiIndex):
+                    # We still have the corner case where a tuple is a key
+                    # in the first level of our MultiIndex
+                    return self._get_values_tuple(key)
 
         if is_iterator(key):
             key = list(key)
@@ -963,7 +961,7 @@ def _get_values_tuple(self, key):
             return result
 
         if not isinstance(self.index, MultiIndex):
-            raise ValueError("Can only tuple-index with a MultiIndex")
+            raise ValueError("key of type tuple not found and not a MultiIndex")
 
         # If key is contained, would have returned by now
         indexer, new_index = self.index.get_loc_level(key)
@@ -1017,9 +1015,11 @@ def __setitem__(self, key, value):
                 # GH#12862 adding an new key to the Series
                 self.loc[key] = value
 
-        except TypeError as e:
+        except TypeError as err:
             if isinstance(key, tuple) and not isinstance(self.index, MultiIndex):
-                raise ValueError("Can only tuple-index with a MultiIndex") from e
+                raise ValueError(
+                    "key of type tuple not found and not a MultiIndex"
+                ) from err
 
             if com.is_bool_indexer(key):
                 key = check_bool_indexer(self.index, key)

pandas/tests/series/indexing/test_indexing.py

@@ -383,7 +383,7 @@ def test_2d_to_1d_assignment_raises():
 @pytest.mark.filterwarnings("ignore:Using a non-tuple:FutureWarning")
 def test_basic_getitem_setitem_corner(datetime_series):
     # invalid tuples, e.g. td.ts[:, None] vs. td.ts[:, 2]
-    msg = "Can only tuple-index with a MultiIndex"
+    msg = "key of type tuple not found and not a MultiIndex"
     with pytest.raises(ValueError, match=msg):
         datetime_series[:, 2]
     with pytest.raises(ValueError, match=msg):
@@ -942,3 +942,22 @@ def assert_slices_equivalent(l_slc, i_slc):
         for key2 in [keystr2, box(keystr2)]:
             assert_slices_equivalent(SLC[key2:key:-1], SLC[13:8:-1])
             assert_slices_equivalent(SLC[key:key2:-1], SLC[0:0:-1])
+
+
+def test_tuple_index():
+    # GH 35534 - Selecting values when a Series has an Index of tuples
+    s = pd.Series([1, 2], index=[("a",), ("b",)])
+    assert s[("a",)] == 1
+    assert s[("b",)] == 2
+    s[("b",)] = 3
+    assert s[("b",)] == 3
+
+
+def test_frozenset_index():
+    # GH35747 - Selecting values when a Series has an Index of frozenset
+    idx0, idx1 = frozenset("a"), frozenset("b")
+    s = pd.Series([1, 2], index=[idx0, idx1])
+    assert s[idx0] == 1
+    assert s[idx1] == 2
+    s[idx1] = 3
+    assert s[idx1] == 3