Bug in xs raising KeyError for MultiIndex columns with droplevel Fals…

…e and list indexe (#41789)

doc/source/whatsnew/v1.2.5.rst

@@ -21,6 +21,13 @@ Fixed regressions
 
 .. ---------------------------------------------------------------------------
 
+.. _whatsnew_125.deprecations:
+
+Deprecations
+~~~~~~~~~~~~
+
+- Deprecated passing lists as ``key`` to :meth:`DataFrame.xs` and :meth:`Series.xs` (:issue:`41760`)
+
 .. _whatsnew_125.bug_fixes:
 
 Bug fixes

pandas/core/generic.py

@@ -3756,6 +3756,15 @@ class   animal   locomotion
         """
         axis = self._get_axis_number(axis)
         labels = self._get_axis(axis)
+
+        if isinstance(key, list):
+            warnings.warn(
+                "Passing lists as key for xs is deprecated and will be removed in a "
+                "future version. Pass key as a tuple instead.",
+                FutureWarning,
+                stacklevel=2,
+            )
+
         if level is not None:
             if not isinstance(labels, MultiIndex):
                 raise TypeError("Index must be a MultiIndex")

pandas/tests/frame/indexing/test_xs.py

@@ -106,7 +106,8 @@ def test_xs_keep_level(self):
         expected = df[:1]
         tm.assert_frame_equal(result, expected)
 
-        result = df.xs([2008, "sat"], level=["year", "day"], drop_level=False)
+        with tm.assert_produces_warning(FutureWarning):
+            result = df.xs([2008, "sat"], level=["year", "day"], drop_level=False)
         tm.assert_frame_equal(result, expected)
 
     def test_xs_view(self, using_array_manager):
@@ -187,7 +188,11 @@ def test_xs_with_duplicates(self, key, level, multiindex_dataframe_random_data):
         assert df.index.is_unique is False
         expected = concat([frame.xs("one", level="second")] * 2)
 
-        result = df.xs(key, level=level)
+        if isinstance(key, list):
+            with tm.assert_produces_warning(FutureWarning):
+                result = df.xs(key, level=level)
+        else:
+            result = df.xs(key, level=level)
         tm.assert_frame_equal(result, expected)
 
     def test_xs_missing_values_in_index(self):
@@ -358,3 +363,11 @@ def test_xs_droplevel_false_view(self, using_array_manager):
         df.iloc[0, 0] = 2
         expected = DataFrame({"a": [1]})
         tm.assert_frame_equal(result, expected)
+
+    def test_xs_list_indexer_droplevel_false(self):
+        # GH#41760
+        mi = MultiIndex.from_tuples([("x", "m", "a"), ("x", "n", "b"), ("y", "o", "c")])
+        df = DataFrame([[1, 2, 3], [4, 5, 6]], columns=mi)
+        with tm.assert_produces_warning(FutureWarning):
+            with pytest.raises(KeyError, match="y"):
+                df.xs(["x", "y"], drop_level=False, axis=1)

pandas/tests/indexing/multiindex/test_partial.py

@@ -69,7 +69,8 @@ def test_xs_partial(
         )
         df = DataFrame(np.random.randn(8, 4), index=index, columns=list("abcd"))
 
-        result = df.xs(["foo", "one"])
+        with tm.assert_produces_warning(FutureWarning):
+            result = df.xs(["foo", "one"])
         expected = df.loc["foo", "one"]
         tm.assert_frame_equal(result, expected)
 

pandas/tests/series/indexing/test_xs.py

@@ -69,3 +69,13 @@ def test_series_xs_droplevel_false(self):
             ),
         )
         tm.assert_series_equal(result, expected)
+
+    def test_xs_key_as_list(self):
+        # GH#41760
+        mi = MultiIndex.from_tuples([("a", "x")], names=["level1", "level2"])
+        ser = Series([1], index=mi)
+        with tm.assert_produces_warning(FutureWarning):
+            ser.xs(["a", "x"], axis=0, drop_level=False)
+
+        with tm.assert_produces_warning(FutureWarning):
+            ser.xs(["a"], axis=0, drop_level=False)