BUG: Fix return type of loc when key is slice

doc/source/whatsnew/v3.0.0.rst

@@ -1217,6 +1217,7 @@ Indexing
 - Bug in :meth:`DataFrame.__setitem__` on an empty :class:`DataFrame` with a tuple corrupting the frame (:issue:`54385`)
 - Bug in :meth:`DataFrame.from_records` throwing a ``ValueError`` when passed an empty list in ``index`` (:issue:`58594`)
 - Bug in :meth:`DataFrame.loc` and :meth:`DataFrame.iloc` returning incorrect dtype when selecting from a :class:`DataFrame` with mixed data types. (:issue:`60600`)
+- Bug in :meth:`DataFrame.loc` returning incorrect dtype when the key is ``slice`` (:issue:`63071`)
 - Bug in :meth:`DataFrame.loc` with inconsistent behavior of loc-set with 2 given indexes to Series (:issue:`59933`)
 - Bug in :meth:`Index.equals` when comparing between :class:`Series` with string dtype :class:`Index` (:issue:`61099`)
 - Bug in :meth:`Index.get_indexer` and similar methods when ``NaN`` is located at or after position 128 (:issue:`58924`)

pandas/core/indexing.py

@@ -1087,7 +1087,11 @@ def _getitem_lowerdim(self, tup: tuple):
         # Reverse tuple so that we are indexing along columns before rows
         # and avoid unintended dtype inference. # GH60600
         for i, key in zip(range(len(tup) - 1, -1, -1), reversed(tup), strict=True):
-            if is_label_like(key) or is_list_like(key):
+            if (
+                is_label_like(key)
+                or is_list_like(key)
+                or (isinstance(key, slice) and need_slice(key))
+            ):
                 # We don't need to check for tuples here because those are
                 #  caught by the _is_nested_tuple_indexer check above.
                 section = self._getitem_axis(key, axis=i)

pandas/tests/indexing/test_loc.py

@@ -61,10 +61,11 @@ def test_not_change_nan_loc(series, new_series, expected_ser):
     tm.assert_frame_equal(df.notna(), ~expected)
 
 
-def test_loc_dtype():
+@pytest.mark.parametrize("key", [[1, 2], slice(1, 3)])
+def test_loc_dtype(key):
     # GH 60600
     df = DataFrame([["a", 1.0, 2.0], ["b", 3.0, 4.0]])
-    result = df.loc[0, [1, 2]]
+    result = df.loc[0, key]
     expected = Series([1.0, 2.0], index=[1, 2], dtype=float, name=0)
     tm.assert_series_equal(result, expected)