Merge branch 'pandas-dev:main' into ordered_cat_corr

Author: pandeconscious

doc/source/whatsnew/v3.0.0.rst

@@ -1300,6 +1300,7 @@ ExtensionArray
 - Bug in :class:`Categorical` when constructing with an :class:`Index` with :class:`ArrowDtype` (:issue:`60563`)
 - Bug in :meth:`.arrays.ArrowExtensionArray.__setitem__` which caused wrong behavior when using an integer array with repeated values as a key (:issue:`58530`)
 - Bug in :meth:`ArrowExtensionArray.factorize` where NA values were dropped when input was dictionary-encoded even when dropna was set to False(:issue:`60567`)
+- Bug in :meth:`NDArrayBackedExtensionArray.take` which produced arrays whose dtypes didn't match their underlying data, when called with integer arrays (:issue:`62448`)
 - Bug in :meth:`api.types.is_datetime64_any_dtype` where a custom :class:`ExtensionDtype` would return ``False`` for array-likes (:issue:`57055`)
 - Bug in comparison between object with :class:`ArrowDtype` and incompatible-dtyped (e.g. string vs bool) incorrectly raising instead of returning all-``False`` (for ``==``) or all-``True`` (for ``!=``) (:issue:`59505`)
 - Bug in constructing pandas data structures when passing into ``dtype`` a string of the type followed by ``[pyarrow]`` while PyArrow is not installed would raise ``NameError`` rather than ``ImportError`` (:issue:`57928`)

pandas/core/arrays/numpy_.py

@@ -48,6 +48,7 @@
         InterpolateOptions,
         NpDtype,
         Scalar,
+        TakeIndexer,
         npt,
     )
 
@@ -365,6 +366,27 @@ def interpolate(
             return self
         return type(self)._simple_new(out_data, dtype=self.dtype)
 
+    def take(
+        self,
+        indices: TakeIndexer,
+        *,
+        allow_fill: bool = False,
+        fill_value: Any = None,
+        axis: AxisInt = 0,
+    ) -> Self:
+        """
+        Take entries from this array at each index in a list of indices,
+        producing an array containing only those entries.
+        """
+        result = super().take(
+            indices, allow_fill=allow_fill, fill_value=fill_value, axis=axis
+        )
+        # See GH#62448.
+        if self.dtype.kind in "iub":
+            return type(self)(result._ndarray, copy=False)
+
+        return result
+
     # ------------------------------------------------------------------------
     # Reductions
 

pandas/tests/arrays/numpy_/test_numpy.py

@@ -348,6 +348,42 @@ def test_factorize_unsigned():
     tm.assert_extension_array_equal(res_unique, NumpyExtensionArray(exp_unique))
 
 
+@pytest.mark.parametrize(
+    "dtype",
+    [
+        np.bool_,
+        np.uint8,
+        np.uint16,
+        np.uint32,
+        np.uint64,
+        np.int8,
+        np.int16,
+        np.int32,
+        np.int64,
+    ],
+)
+def test_take_assigns_floating_point_dtype(dtype):
+    # GH#62448.
+    if dtype == np.bool_:
+        array = NumpyExtensionArray(np.array([False, True, False], dtype=dtype))
+        expected = np.dtype(object)
+    else:
+        array = NumpyExtensionArray(np.array([1, 2, 3], dtype=dtype))
+        expected = np.float64
+
+    result = array.take([-1], allow_fill=True)
+    assert result.dtype.numpy_dtype == expected
+
+    result = array.take([-1], allow_fill=True, fill_value=5.0)
+    assert result.dtype.numpy_dtype == expected
+
+
+def test_take_preserves_boolean_arrays():
+    array = NumpyExtensionArray(np.array([False, True, False], dtype=np.bool_))
+    result = array.take([-1], allow_fill=False)
+    assert result.dtype.numpy_dtype == np.bool_
+
+
 # ----------------------------------------------------------------------------
 # Output formatting
 

pandas/tests/tseries/offsets/test_month.py

@@ -61,7 +61,7 @@ def test_offset_whole_year(self):
             datetime(2008, 12, 31),
         )
 
-        for base, exp_date in zip(dates[:-1], dates[1:]):
+        for base, exp_date in zip(dates[:-1], dates[1:], strict=True):
             assert_offset_equal(SemiMonthEnd(), base, exp_date)
 
         # ensure .apply_index works as expected
@@ -312,7 +312,7 @@ def test_offset_whole_year(self):
             datetime(2008, 12, 15),
         )
 
-        for base, exp_date in zip(dates[:-1], dates[1:]):
+        for base, exp_date in zip(dates[:-1], dates[1:], strict=True):
             assert_offset_equal(SemiMonthBegin(), base, exp_date)
 
         # ensure .apply_index works as expected

pandas/tests/tseries/offsets/test_offsets.py

@@ -649,6 +649,7 @@ def test_copy(self):
                 "2008-01-02 00:00:00.001000000",
                 "2008-01-02 00:00:00.000001000",
             ],
+            strict=True,
         ),
     )
     def test_add(self, arithmatic_offset_type, expected, dt):
@@ -670,6 +671,7 @@ def test_add(self, arithmatic_offset_type, expected, dt):
                 "2008-01-01 23:59:59.999000000",
                 "2008-01-01 23:59:59.999999000",
             ],
+            strict=True,
         ),
     )
     def test_sub(self, arithmatic_offset_type, expected, dt):
@@ -693,6 +695,7 @@ def test_sub(self, arithmatic_offset_type, expected, dt):
                 "2008-01-02 00:00:00.008000000",
                 "2008-01-02 00:00:00.000009000",
             ],
+            strict=True,
         ),
     )
     def test_mul_add(self, arithmatic_offset_type, n, expected, dt):
@@ -717,6 +720,7 @@ def test_mul_add(self, arithmatic_offset_type, n, expected, dt):
                 "2008-01-01 23:59:59.992000000",
                 "2008-01-01 23:59:59.999991000",
             ],
+            strict=True,
         ),
     )
     def test_mul_sub(self, arithmatic_offset_type, n, expected, dt):

pyproject.toml

@@ -588,8 +588,6 @@ exclude = [
 "pandas/tests/strings/test_strings.py" = ["B905"]
 "pandas/tests/test_algos.py" = ["B905"]
 "pandas/tests/test_sorting.py" = ["B905"]
-"pandas/tests/tseries/offsets/test_month.py" = ["B905"]
-"pandas/tests/tseries/offsets/test_offsets.py" = ["B905"]
 "pandas/tests/util/test_validate_kwargs.py" = ["B905"]
 "scripts/validate_unwanted_patterns.py" = ["B905"]