DOC/TST: Document numpy 2.0 support and add tests for string array

doc/source/whatsnew/v2.2.2.rst

@@ -9,6 +9,21 @@ including other versions of pandas.
 {{ header }}
 
 .. ---------------------------------------------------------------------------
+
+.. _whatsnew_220.np2_compat:
+
+Pandas 2.2.2 is now compatible with numpy 2.0
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+Pandas 2.2.2 is the first version of pandas that is generally compatible with the upcoming
+numpy 2.0 release, and wheels for pandas 2.2.2 will work with both numpy 1.x and 2.x.
+
+One major caveat is that arrays created with numpy 2.0's new ``StringDtype`` will convert
+to ``object`` dtyped arrays upon :class:`Series`/:class:`DataFrame` creation.
+Full support for numpy 2.0's StringDtype is expected to land in pandas 3.0.
+
+As usual please report any bugs discovered to our `issue tracker <https://github.com/pandas-dev/pandas/issues/new/choose>`_
+
 .. _whatsnew_222.regressions:
 
 Fixed regressions

pandas/tests/frame/test_constructors.py

@@ -24,6 +24,7 @@
 from pandas._config import using_pyarrow_string_dtype
 
 from pandas._libs import lib
+from pandas.compat.numpy import np_version_gt2
 from pandas.errors import IntCastingNaNError
 
 from pandas.core.dtypes.common import is_integer_dtype
@@ -3052,6 +3053,24 @@ def test_from_dict_with_columns_na_scalar(self):
         expected = DataFrame({"a": Series([pd.NaT, pd.NaT])})
         tm.assert_frame_equal(result, expected)
 
+    # TODO: make this not cast to object in pandas 3.0
+    @pytest.mark.skipif(
+        not np_version_gt2, reason="StringDType only available in numpy 2 and above"
+    )
+    @pytest.mark.parametrize(
+        "data",
+        [
+            {"a": ["a", "b", "c"], "b": [1.0, 2.0, 3.0], "c": ["d", "e", "f"]},
+        ],
+    )
+    def test_np_string_array_object_cast(self, data):
+        from numpy.dtypes import StringDType
+
+        data["a"] = np.array(data["a"], dtype=StringDType())
+        res = DataFrame(data)
+        assert res["a"].dtype == np.object_
+        assert (res["a"] == data["a"]).all()
+
 
 def get1(obj):  # TODO: make a helper in tm?
     if isinstance(obj, Series):

pandas/tests/series/test_constructors.py

@@ -2176,6 +2176,25 @@ def test_series_constructor_infer_multiindex(self, container, data):
         multi = Series(data, index=indexes)
         assert isinstance(multi.index, MultiIndex)
 
+    # TODO: make this not cast to object in pandas 3.0
+    @pytest.mark.skipif(
+        not np_version_gt2, reason="StringDType only available in numpy 2 and above"
+    )
+    @pytest.mark.parametrize(
+        "data",
+        [
+            ["a", "b", "c"],
+            ["a", "b", np.nan],
+        ],
+    )
+    def test_np_string_array_object_cast(self, data):
+        from numpy.dtypes import StringDType
+
+        arr = np.array(data, dtype=StringDType())
+        res = Series(arr)
+        assert res.dtype == np.object_
+        assert (res == data).all()
+
 
 class TestSeriesConstructorInternals:
     def test_constructor_no_pandas_array(self):