TST: tighten PandasArray tests (#39370)

Author: jbrockmendel

pandas/tests/extension/test_numpy.py

@@ -16,6 +16,8 @@
 import numpy as np
 import pytest
 
+from pandas.core.dtypes.dtypes import ExtensionDtype
+
 import pandas as pd
 import pandas._testing as tm
 from pandas.core.arrays.numpy_ import PandasArray, PandasDtype
@@ -121,7 +123,7 @@ def data_for_grouping(allow_in_pandas, dtype):
 
 
 @pytest.fixture
-def skip_numpy_object(dtype):
+def skip_numpy_object(dtype, request):
     """
     Tests for PandasArray with nested data. Users typically won't create
     these objects via `pd.array`, but they can show up through `.array`
@@ -132,14 +134,25 @@ def skip_numpy_object(dtype):
     marker to either an individual test or a test class.
     """
     if dtype == "object":
-        raise pytest.skip("Skipping for object dtype.")
+        mark = pytest.mark.xfail(reason="Fails for object dtype")
+        request.node.add_marker(mark)
 
 
 skip_nested = pytest.mark.usefixtures("skip_numpy_object")
 
 
 class BaseNumPyTests:
-    pass
+    @classmethod
+    def assert_series_equal(cls, left, right, *args, **kwargs):
+        # base class tests hard-code expected values with numpy dtypes,
+        #  whereas we generally want the corresponding PandasDtype
+        if (
+            isinstance(right, pd.Series)
+            and not isinstance(right.dtype, ExtensionDtype)
+            and isinstance(left.dtype, PandasDtype)
+        ):
+            right = right.astype(PandasDtype(right.dtype))
+        return tm.assert_series_equal(left, right, *args, **kwargs)
 
 
 class TestCasting(BaseNumPyTests, base.BaseCastingTests):
@@ -148,24 +161,13 @@ def test_astype_str(self, data):
         # ValueError: setting an array element with a sequence
         super().test_astype_str(data)
 
-    @skip_nested
-    def test_astype_string(self, data):
-        # GH-33465
-        # ValueError: setting an array element with a sequence
-        super().test_astype_string(data)
-
 
 class TestConstructors(BaseNumPyTests, base.BaseConstructorsTests):
     @pytest.mark.skip(reason="We don't register our dtype")
     # We don't want to register. This test should probably be split in two.
     def test_from_dtype(self, data):
         pass
 
-    @skip_nested
-    def test_array_from_scalars(self, data):
-        # ValueError: PandasArray must be 1-dimensional.
-        super().test_array_from_scalars(data)
-
     @skip_nested
     def test_series_constructor_scalar_with_index(self, data, dtype):
         # ValueError: Length of passed values is 1, index implies 3.
@@ -202,10 +204,16 @@ def test_loc_iloc_frame_single_dtype(self, data, request):
 
 
 class TestGroupby(BaseNumPyTests, base.BaseGroupbyTests):
-    @skip_nested
     def test_groupby_extension_apply(
         self, data_for_grouping, groupby_apply_op, request
     ):
+        dummy = groupby_apply_op([None])
+        if (
+            isinstance(dummy, pd.Series)
+            and data_for_grouping.dtype.numpy_dtype == object
+        ):
+            mark = pytest.mark.xfail(reason="raises in MultiIndex construction")
+            request.node.add_marker(mark)
         super().test_groupby_extension_apply(data_for_grouping, groupby_apply_op)
 
 
@@ -225,17 +233,6 @@ def test_value_counts(self, all_data, dropna):
     def test_value_counts_with_normalize(self, data):
         return super().test_value_counts_with_normalize(data)
 
-    @pytest.mark.skip(reason="Incorrect expected")
-    # We have a bool dtype, so the result is an ExtensionArray
-    # but expected is not
-    def test_combine_le(self, data_repeated):
-        super().test_combine_le(data_repeated)
-
-    @skip_nested
-    def test_combine_add(self, data_repeated):
-        # Not numeric
-        super().test_combine_add(data_repeated)
-
     @skip_nested
     def test_shift_fill_value(self, data):
         # np.array shape inference. Shift implementation fails.
@@ -258,11 +255,6 @@ def test_fillna_copy_series(self, data_missing):
         # The "scalar" for this array isn't a scalar.
         super().test_fillna_copy_series(data_missing)
 
-    @skip_nested
-    def test_hash_pandas_object_works(self, data, as_frame):
-        # ndarray of tuples not hashable
-        super().test_hash_pandas_object_works(data, as_frame)
-
     @skip_nested
     def test_searchsorted(self, data_for_sorting, as_series):
         # Test setup fails.
@@ -273,41 +265,51 @@ def test_where_series(self, data, na_value, as_frame):
         # Test setup fails.
         super().test_where_series(data, na_value, as_frame)
 
-    @skip_nested
     @pytest.mark.parametrize("repeats", [0, 1, 2, [1, 2, 3]])
-    def test_repeat(self, data, repeats, as_series, use_numpy):
-        # Fails creating expected
+    def test_repeat(self, data, repeats, as_series, use_numpy, request):
+        if data.dtype.numpy_dtype == object and repeats != 0:
+            mark = pytest.mark.xfail(reason="mask shapes mismatch")
+            request.node.add_marker(mark)
         super().test_repeat(data, repeats, as_series, use_numpy)
 
     @pytest.mark.xfail(reason="PandasArray.diff may fail on dtype")
     def test_diff(self, data, periods):
         return super().test_diff(data, periods)
 
-    @skip_nested
     @pytest.mark.parametrize("box", [pd.array, pd.Series, pd.DataFrame])
-    def test_equals(self, data, na_value, as_series, box):
+    def test_equals(self, data, na_value, as_series, box, request):
         # Fails creating with _from_sequence
+        if box is pd.DataFrame and data.dtype.numpy_dtype == object:
+            mark = pytest.mark.xfail(reason="AssertionError in _get_same_shape_values")
+            request.node.add_marker(mark)
+
         super().test_equals(data, na_value, as_series, box)
 
 
-@skip_nested
 class TestArithmetics(BaseNumPyTests, base.BaseArithmeticOpsTests):
     divmod_exc = None
     series_scalar_exc = None
     frame_scalar_exc = None
     series_array_exc = None
 
+    @skip_nested
+    def test_divmod(self, data):
+        super().test_divmod(data)
+
+    @skip_nested
     def test_divmod_series_array(self, data):
-        s = pd.Series(data)
-        self._check_divmod_op(s, divmod, data, exc=None)
+        ser = pd.Series(data)
+        self._check_divmod_op(ser, divmod, data, exc=None)
 
     @pytest.mark.skip("We implement ops")
     def test_error(self, data, all_arithmetic_operators):
         pass
 
+    @skip_nested
     def test_arith_series_with_scalar(self, data, all_arithmetic_operators):
         super().test_arith_series_with_scalar(data, all_arithmetic_operators)
 
+    @skip_nested
     def test_arith_series_with_array(self, data, all_arithmetic_operators):
         super().test_arith_series_with_array(data, all_arithmetic_operators)
 
@@ -316,14 +318,17 @@ class TestPrinting(BaseNumPyTests, base.BasePrintingTests):
     pass
 
 
-@skip_nested
 class TestNumericReduce(BaseNumPyTests, base.BaseNumericReduceTests):
     def check_reduce(self, s, op_name, skipna):
         result = getattr(s, op_name)(skipna=skipna)
         # avoid coercing int -> float. Just cast to the actual numpy type.
         expected = getattr(s.astype(s.dtype._dtype), op_name)(skipna=skipna)
         tm.assert_almost_equal(result, expected)
 
+    @pytest.mark.parametrize("skipna", [True, False])
+    def test_reduce_series(self, data, all_boolean_reductions, skipna):
+        super().test_reduce_series(data, all_boolean_reductions, skipna)
+
 
 @skip_nested
 class TestBooleanReduce(BaseNumPyTests, base.BaseBooleanReduceTests):
@@ -364,11 +369,6 @@ def test_fillna_fill_other(self, data_missing):
 
 
 class TestReshaping(BaseNumPyTests, base.BaseReshapingTests):
-    @pytest.mark.skip("Incorrect parent test")
-    # not actually a mixed concat, since we concat int and int.
-    def test_concat_mixed_dtypes(self, data):
-        super().test_concat_mixed_dtypes(data)
-
     @skip_nested
     def test_merge(self, data, na_value):
         # Fails creating expected
@@ -390,22 +390,6 @@ def test_transpose_frame(self, data):
 
 
 class TestSetitem(BaseNumPyTests, base.BaseSetitemTests):
-    @skip_nested
-    def test_setitem_scalar_series(self, data, box_in_series):
-        # AssertionError
-        super().test_setitem_scalar_series(data, box_in_series)
-
-    @skip_nested
-    def test_setitem_sequence(self, data, box_in_series):
-        # ValueError: shape mismatch: value array of shape (2,1) could not
-        # be broadcast to indexing result of shape (2,)
-        super().test_setitem_sequence(data, box_in_series)
-
-    @skip_nested
-    def test_setitem_sequence_mismatched_length_raises(self, data, as_array):
-        # ValueError: PandasArray must be 1-dimensional.
-        super().test_setitem_sequence_mismatched_length_raises(data, as_array)
-
     @skip_nested
     def test_setitem_sequence_broadcasts(self, data, box_in_series):
         # ValueError: cannot set using a list-like indexer with a different
@@ -459,7 +443,6 @@ def test_setitem_scalar_key_sequence_raise(self, data):
     def test_setitem_mask(self, data, mask, box_in_series):
         super().test_setitem_mask(data, mask, box_in_series)
 
-    @skip_nested
     def test_setitem_mask_raises(self, data, box_in_series):
         super().test_setitem_mask_raises(data, box_in_series)
 
@@ -472,7 +455,6 @@ def test_setitem_mask_raises(self, data, box_in_series):
     def test_setitem_integer_array(self, data, idx, box_in_series):
         super().test_setitem_integer_array(data, idx, box_in_series)
 
-    @skip_nested
     @pytest.mark.parametrize(
         "idx, box_in_series",
         [