TST: Simplify and clarify test skipping (#56027)

* TST: Simplify and clarify test skipping

* Typo

* Update pandas/tests/arithmetic/test_datetime64.py

* Don't skip because of duplicates

Author: mroeschke

pandas/tests/arithmetic/test_datetime64.py

@@ -414,7 +414,7 @@ def test_dti_cmp_datetimelike(self, other, tz_naive_fixture):
         dti = date_range("2016-01-01", periods=2, tz=tz)
         if tz is not None:
             if isinstance(other, np.datetime64):
-                pytest.skip("no tzaware version available")
+                pytest.skip(f"{type(other).__name__} is not tz aware")
             other = localize_pydatetime(other, dti.tzinfo)
 
         result = dti == other

pandas/tests/arithmetic/test_numeric.py

@@ -384,7 +384,7 @@ def test_divmod_zero(self, zero, numeric_idx):
     def test_div_negative_zero(self, zero, numeric_idx, op):
         # Check that -1 / -0.0 returns np.inf, not -np.inf
         if numeric_idx.dtype == np.uint64:
-            pytest.skip(f"Not relevant for {numeric_idx.dtype}")
+            pytest.skip(f"Div by negative 0 not relevant for {numeric_idx.dtype}")
         idx = numeric_idx - 3
 
         expected = Index([-np.inf, -np.inf, -np.inf, np.nan, np.inf], dtype=np.float64)

pandas/tests/computation/test_eval.py

@@ -63,7 +63,7 @@
                     reason=f"numexpr enabled->{USE_NUMEXPR}, "
                     f"installed->{NUMEXPR_INSTALLED}",
                 ),
-                td.skip_if_no_ne,
+                td.skip_if_no("numexpr"),
             ],
         )
         for engine in ENGINES
@@ -1687,14 +1687,14 @@ def test_empty_globals(self, engine, parser):
             pd.eval(e, engine=engine, parser=parser, global_dict={})
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 def test_invalid_engine():
     msg = "Invalid engine 'asdf' passed"
     with pytest.raises(KeyError, match=msg):
         pd.eval("x + y", local_dict={"x": 1, "y": 2}, engine="asdf")
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 @pytest.mark.parametrize(
     ("use_numexpr", "expected"),
     (
@@ -1711,7 +1711,7 @@ def test_numexpr_option_respected(use_numexpr, expected):
         assert result == expected
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 def test_numexpr_option_incompatible_op():
     # GH 32556
     with pd.option_context("compute.use_numexpr", False):
@@ -1723,7 +1723,7 @@ def test_numexpr_option_incompatible_op():
         tm.assert_frame_equal(result, expected)
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 def test_invalid_parser():
     msg = "Invalid parser 'asdf' passed"
     with pytest.raises(KeyError, match=msg):

pandas/tests/dtypes/test_common.py

@@ -196,7 +196,7 @@ def test_is_object():
 
 
 @pytest.mark.parametrize(
-    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
 )
 def test_is_sparse(check_scipy):
     msg = "is_sparse is deprecated"
@@ -632,7 +632,7 @@ def test_is_bool_dtype_numpy_error():
 
 
 @pytest.mark.parametrize(
-    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+    "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
 )
 def test_is_extension_array_dtype(check_scipy):
     assert not com.is_extension_array_dtype([1, 2, 3])

pandas/tests/extension/base/dim2.py

@@ -29,7 +29,7 @@ def skip_if_doesnt_support_2d(self, dtype, request):
             test_func = node._obj
             if test_func.__qualname__.startswith("Dim2CompatTests"):
                 # TODO: is there a less hacky way of checking this?
-                pytest.skip("Test is only for EAs that support 2D.")
+                pytest.skip(f"{dtype} does not support 2D.")
 
     def test_transpose(self, data):
         arr2d = data.repeat(2).reshape(-1, 2)

pandas/tests/extension/base/interface.py

@@ -103,7 +103,7 @@ def test_copy(self, data):
         result = data.copy()
 
         if data.dtype._is_immutable:
-            pytest.skip("test_copy assumes mutability")
+            pytest.skip(f"test_copy assumes mutability and {data.dtype} is immutable")
 
         data[1] = data[0]
         assert result[1] != result[0]
@@ -118,7 +118,7 @@ def test_view(self, data):
         assert type(result) == type(data)
 
         if data.dtype._is_immutable:
-            pytest.skip("test_view assumes mutability")
+            pytest.skip(f"test_view assumes mutability and {data.dtype} is immutable")
 
         result[1] = result[0]
         assert data[1] == data[0]

pandas/tests/extension/base/reduce.py

@@ -120,7 +120,7 @@ def test_reduce_frame(self, data, all_numeric_reductions, skipna):
         op_name = all_numeric_reductions
         ser = pd.Series(data)
         if not is_numeric_dtype(ser.dtype):
-            pytest.skip("not numeric dtype")
+            pytest.skip(f"{ser.dtype} is not numeric dtype")
 
         if op_name in ["count", "kurt", "sem"]:
             pytest.skip(f"{op_name} not an array method")

pandas/tests/extension/base/reshaping.py

@@ -343,7 +343,7 @@ def test_ravel(self, data):
         assert type(result) == type(data)
 
         if data.dtype._is_immutable:
-            pytest.skip("test_ravel assumes mutability")
+            pytest.skip(f"test_ravel assumes mutability and {data.dtype} is immutable")
 
         # Check that we have a view, not a copy
         result[0] = result[1]
@@ -360,7 +360,9 @@ def test_transpose(self, data):
         assert result.shape == data.shape[::-1]
 
         if data.dtype._is_immutable:
-            pytest.skip("test_transpose assumes mutability")
+            pytest.skip(
+                f"test_transpose assumes mutability and {data.dtype} is immutable"
+            )
 
         # Check that we have a view, not a copy
         result[0] = result[1]

pandas/tests/extension/base/setitem.py

@@ -43,7 +43,7 @@ def skip_if_immutable(self, dtype, request):
                 # This fixture is auto-used, but we want to not-skip
                 # test_is_immutable.
                 return
-            pytest.skip("__setitem__ test not applicable with immutable dtype")
+            pytest.skip(f"__setitem__ test not applicable with immutable dtype {dtype}")
 
     def test_is_immutable(self, data):
         if data.dtype._is_immutable:

pandas/tests/extension/conftest.py

@@ -37,7 +37,7 @@ def data_for_twos(dtype):
     if not (dtype._is_numeric or dtype.kind == "m"):
         # Object-dtypes may want to allow this, but for the most part
         #  only numeric and timedelta-like dtypes will need to implement this.
-        pytest.skip("Not a numeric dtype")
+        pytest.skip(f"{dtype} is not a numeric dtype")
 
     raise NotImplementedError
 

pandas/tests/extension/test_interval.py

@@ -56,7 +56,7 @@ def data_missing():
 
 @pytest.fixture
 def data_for_twos():
-    pytest.skip("Not a numeric dtype")
+    pytest.skip("Interval is not a numeric dtype")
 
 
 @pytest.fixture

pandas/tests/extension/test_numpy.py

@@ -151,7 +151,7 @@ def data_for_grouping(allow_in_pandas, dtype):
 @pytest.fixture
 def data_for_twos(dtype):
     if dtype.kind == "O":
-        pytest.skip("Not a numeric dtype")
+        pytest.skip(f"{dtype} is not a numeric dtype")
     arr = np.ones(100) * 2
     return NumpyExtensionArray._from_sequence(arr, dtype=dtype)
 
@@ -323,7 +323,7 @@ def check_reduce(self, ser: pd.Series, op_name: str, skipna: bool):
             expected = exp_op(skipna=skipna)
         tm.assert_almost_equal(result, expected)
 
-    @pytest.mark.skip("tests not written yet")
+    @pytest.mark.skip("TODO: tests not written yet")
     @pytest.mark.parametrize("skipna", [True, False])
     def test_reduce_frame(self, data, all_numeric_reductions, skipna):
         pass

pandas/tests/extension/test_string.py

@@ -26,22 +26,21 @@
 from pandas.tests.extension import base
 
 
-def split_array(arr):
-    if arr.dtype.storage != "pyarrow":
-        pytest.skip("only applicable for pyarrow chunked array n/a")
-
-    def _split_array(arr):
-        import pyarrow as pa
-
-        arrow_array = arr._pa_array
-        split = len(arrow_array) // 2
-        arrow_array = pa.chunked_array(
-            [*arrow_array[:split].chunks, *arrow_array[split:].chunks]
-        )
-        assert arrow_array.num_chunks == 2
-        return type(arr)(arrow_array)
-
-    return _split_array(arr)
+def maybe_split_array(arr, chunked):
+    if not chunked:
+        return arr
+    elif arr.dtype.storage != "pyarrow":
+        return arr
+
+    pa = pytest.importorskip("pyarrow")
+
+    arrow_array = arr._pa_array
+    split = len(arrow_array) // 2
+    arrow_array = pa.chunked_array(
+        [*arrow_array[:split].chunks, *arrow_array[split:].chunks]
+    )
+    assert arrow_array.num_chunks == 2
+    return type(arr)(arrow_array)
 
 
 @pytest.fixture(params=[True, False])
@@ -61,34 +60,34 @@ def data(dtype, chunked):
         strings = np.random.default_rng(2).choice(list(string.ascii_letters), size=100)
 
     arr = dtype.construct_array_type()._from_sequence(strings)
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_missing(dtype, chunked):
     """Length 2 array with [NA, Valid]"""
     arr = dtype.construct_array_type()._from_sequence([pd.NA, "A"])
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_for_sorting(dtype, chunked):
     arr = dtype.construct_array_type()._from_sequence(["B", "C", "A"])
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_missing_for_sorting(dtype, chunked):
     arr = dtype.construct_array_type()._from_sequence(["B", pd.NA, "A"])
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 @pytest.fixture
 def data_for_grouping(dtype, chunked):
     arr = dtype.construct_array_type()._from_sequence(
         ["B", "B", pd.NA, pd.NA, "A", "A", "B", "C"]
     )
-    return split_array(arr) if chunked else arr
+    return maybe_split_array(arr, chunked)
 
 
 class TestDtype(base.BaseDtypeTests):

pandas/tests/frame/methods/test_interpolate.py

@@ -322,7 +322,7 @@ def test_rowwise_alt(self):
         # TODO: assert something?
 
     @pytest.mark.parametrize(
-        "check_scipy", [False, pytest.param(True, marks=td.skip_if_no_scipy)]
+        "check_scipy", [False, pytest.param(True, marks=td.skip_if_no("scipy"))]
     )
     def test_interp_leading_nans(self, check_scipy):
         df = DataFrame(

pandas/tests/frame/test_query_eval.py

@@ -27,15 +27,16 @@ def parser(request):
 
 
 @pytest.fixture(
-    params=["python", pytest.param("numexpr", marks=td.skip_if_no_ne)], ids=lambda x: x
+    params=["python", pytest.param("numexpr", marks=td.skip_if_no("numexpr"))],
+    ids=lambda x: x,
 )
 def engine(request):
     return request.param
 
 
 def skip_if_no_pandas_parser(parser):
     if parser != "pandas":
-        pytest.skip(f"cannot evaluate with parser {repr(parser)}")
+        pytest.skip(f"cannot evaluate with parser={parser}")
 
 
 class TestCompat:
@@ -387,7 +388,7 @@ def to_series(mi, level):
                 raise AssertionError("object must be a Series or Index")
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 class TestDataFrameQueryNumExprPandas:
     @pytest.fixture
     def engine(self):
@@ -765,7 +766,7 @@ def test_method_calls_in_query(self, engine, parser):
         tm.assert_frame_equal(result, expected)
 
 
-@td.skip_if_no_ne
+@td.skip_if_no("numexpr")
 class TestDataFrameQueryNumExprPython(TestDataFrameQueryNumExprPandas):
     @pytest.fixture
     def engine(self):

pandas/tests/frame/test_reductions.py

@@ -165,8 +165,8 @@ class TestDataFrameAnalytics:
             "var",
             "std",
             "sem",
-            pytest.param("skew", marks=td.skip_if_no_scipy),
-            pytest.param("kurt", marks=td.skip_if_no_scipy),
+            pytest.param("skew", marks=td.skip_if_no("scipy")),
+            pytest.param("kurt", marks=td.skip_if_no("scipy")),
         ],
     )
     def test_stat_op_api_float_string_frame(self, float_string_frame, axis, opname):
@@ -219,8 +219,8 @@ def test_stat_op_api_float_string_frame(self, float_string_frame, axis, opname):
             "var",
             "std",
             "sem",
-            pytest.param("skew", marks=td.skip_if_no_scipy),
-            pytest.param("kurt", marks=td.skip_if_no_scipy),
+            pytest.param("skew", marks=td.skip_if_no("scipy")),
+            pytest.param("kurt", marks=td.skip_if_no("scipy")),
         ],
     )
     def test_stat_op_api_float_frame(self, float_frame, axis, opname):

pandas/tests/groupby/test_reductions.py

@@ -933,7 +933,7 @@ def scipy_sem(*args, **kwargs):
         ("first", lambda x: x.iloc[0]),
         ("last", lambda x: x.iloc[-1]),
         ("count", np.size),
-        pytest.param("sem", scipy_sem, marks=td.skip_if_no_scipy),
+        pytest.param("sem", scipy_sem, marks=td.skip_if_no("scipy")),
     ],
 )
 def test_ops_general(op, targop):

pandas/tests/indexes/test_setops.py

@@ -320,8 +320,9 @@ def test_corner_union(self, index_flat, fname, sname, expected_name):
         # Test unions with various name combinations
         # Do not test MultiIndex or repeats
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # Test copy.union(copy)
         first = index.copy().set_names(fname)
@@ -363,8 +364,9 @@ def test_corner_union(self, index_flat, fname, sname, expected_name):
     )
     def test_union_unequal(self, index_flat, fname, sname, expected_name):
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # test copy.union(subset) - need sort for unicode and string
         first = index.copy().set_names(fname)
@@ -387,8 +389,9 @@ def test_corner_intersect(self, index_flat, fname, sname, expected_name):
         # GH#35847
         # Test intersections with various name combinations
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # Test copy.intersection(copy)
         first = index.copy().set_names(fname)
@@ -430,8 +433,9 @@ def test_corner_intersect(self, index_flat, fname, sname, expected_name):
     )
     def test_intersect_unequal(self, index_flat, fname, sname, expected_name):
         if not index_flat.is_unique:
-            pytest.skip("Randomly generated index_flat was not unique.")
-        index = index_flat
+            index = index_flat.unique()
+        else:
+            index = index_flat
 
         # test copy.intersection(subset) - need sort for unicode and string
         first = index.copy().set_names(fname)