CI: Skip numpy dev failing tests (#39090)

* CI: Skip numpy dev failing tests

* Xfail missed test

* Pull is_numpy_dev up

Author: phofl

pandas/compat/__init__.py

@@ -12,6 +12,7 @@
 import warnings
 
 from pandas._typing import F
+from pandas.compat.numpy import is_numpy_dev
 
 PY38 = sys.version_info >= (3, 8)
 PY39 = sys.version_info >= (3, 9)
@@ -118,3 +119,8 @@ def get_lzma_file(lzma):
             "might be required to solve this issue."
         )
     return lzma.LZMAFile
+
+
+__all__ = [
+    "is_numpy_dev",
+]

pandas/tests/arithmetic/test_interval.py

@@ -3,6 +3,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 from pandas.core.dtypes.common import is_list_like
 
 import pandas as pd
@@ -252,6 +254,7 @@ def test_compare_length_mismatch_errors(self, op, other_constructor, length):
         with pytest.raises(ValueError, match="Lengths must match to compare"):
             op(array, other)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     @pytest.mark.parametrize(
         "constructor, expected_type, assert_func",
         [

pandas/tests/base/test_misc.py

@@ -3,7 +3,7 @@
 import numpy as np
 import pytest
 
-from pandas.compat import IS64, PYPY
+from pandas.compat import IS64, PYPY, is_numpy_dev
 
 from pandas.core.dtypes.common import is_categorical_dtype, is_object_dtype
 
@@ -116,6 +116,9 @@ def test_searchsorted(index_or_series_obj):
         # See gh-14833
         pytest.skip("np.searchsorted doesn't work on pd.MultiIndex")
 
+    if is_object_dtype(obj) and is_numpy_dev:
+        pytest.skip("GH#39089 Numpy changed dtype inference")
+
     max_obj = max(obj, default=0)
     index = np.searchsorted(obj, max_obj)
     assert 0 <= index <= len(obj)

pandas/tests/extension/base/methods.py

@@ -3,6 +3,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 from pandas.core.dtypes.common import is_bool_dtype
 
 import pandas as pd
@@ -392,6 +394,9 @@ def test_hash_pandas_object_works(self, data, as_frame):
         b = pd.util.hash_pandas_object(data)
         self.assert_equal(a, b)
 
+    @pytest.mark.xfail(
+        is_numpy_dev, reason="GH#39089 Numpy changed dtype inference", strict=False
+    )
     def test_searchsorted(self, data_for_sorting, as_series):
         b, c, a = data_for_sorting
         arr = type(data_for_sorting)._from_sequence([a, b, c])

pandas/tests/frame/apply/test_frame_apply.py

@@ -5,6 +5,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 from pandas.core.dtypes.dtypes import CategoricalDtype
 
 import pandas as pd
@@ -582,6 +584,7 @@ def test_apply_dict(self):
             tm.assert_frame_equal(reduce_false, df)
             tm.assert_series_equal(reduce_none, dicts)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_applymap(self, float_frame):
         applied = float_frame.applymap(lambda x: x * 2)
         tm.assert_frame_equal(applied, float_frame * 2)

pandas/tests/frame/indexing/test_indexing.py

@@ -5,6 +5,7 @@
 import pytest
 
 from pandas._libs import iNaT
+from pandas.compat import is_numpy_dev
 
 from pandas.core.dtypes.common import is_integer
 
@@ -254,6 +255,8 @@ def inc(x):
     )
     def test_setitem_same_column(self, cols, values, expected):
         # GH 23239
+        if cols == ["C", "D", "D", "a"] and is_numpy_dev:
+            pytest.skip("GH#39089 Numpy changed dtype inference")
         df = DataFrame([values], columns=cols)
         df["a"] = df["a"]
         result = df["a"].values[0]

pandas/tests/frame/methods/test_drop.py

@@ -3,6 +3,7 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
 from pandas.errors import PerformanceWarning
 
 import pandas as pd
@@ -107,6 +108,7 @@ def test_drop_names(self):
         expected = Index(["a", "b", "c"], name="first")
         tm.assert_index_equal(dropped.index, expected)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_drop(self):
         simple = DataFrame({"A": [1, 2, 3, 4], "B": [0, 1, 2, 3]})
         tm.assert_frame_equal(simple.drop("A", axis=1), simple[["B"]])

pandas/tests/frame/methods/test_isin.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 import pandas as pd
 from pandas import DataFrame, MultiIndex, Series
 import pandas._testing as tm
@@ -32,6 +34,7 @@ def test_isin_empty(self, empty):
         result = df.isin(empty)
         tm.assert_frame_equal(result, expected)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_isin_dict(self):
         df = DataFrame({"A": ["a", "b", "c"], "B": ["a", "e", "f"]})
         d = {"A": ["a"]}

pandas/tests/frame/methods/test_replace.py

@@ -6,6 +6,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 import pandas as pd
 from pandas import DataFrame, Index, Series, Timestamp, date_range
 import pandas._testing as tm
@@ -1508,6 +1510,7 @@ def test_replace_no_replacement_dtypes(self, dtype, value):
         result = df.replace(to_replace=[None, -np.inf, np.inf], value=value)
         tm.assert_frame_equal(result, df)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     @pytest.mark.parametrize("replacement", [np.nan, 5])
     def test_replace_with_duplicate_columns(self, replacement):
         # GH 24798

pandas/tests/frame/methods/test_to_csv.py

@@ -5,6 +5,7 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
 from pandas.errors import ParserError
 
 import pandas as pd
@@ -181,6 +182,7 @@ def test_to_csv_cols_reordering(self):
 
         tm.assert_frame_equal(df[cols], rs_c, check_names=False)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_to_csv_new_dupe_cols(self):
         import pandas as pd
 

pandas/tests/frame/test_nonunique_indexes.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 import pandas as pd
 from pandas import DataFrame, MultiIndex, Series, date_range
 import pandas._testing as tm
@@ -14,6 +16,7 @@ def check(result, expected=None):
 
 
 class TestDataFrameNonuniqueIndexes:
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_column_dups_operations(self):
 
         # assignment
@@ -310,6 +313,7 @@ def test_column_dups2(self):
         result = df.dropna(subset=["A", "C"], how="all")
         tm.assert_frame_equal(result, expected)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_getitem_boolean_series_with_duplicate_columns(self):
         # boolean indexing
         # GH 4879
@@ -337,6 +341,7 @@ def test_getitem_boolean_frame_with_duplicate_columns(self):
         result = df[df > 6]
         check(result, expected)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_getitem_boolean_frame_unaligned_with_duplicate_columns(self):
         # `df.A > 6` is a DataFrame with a different shape from df
         dups = ["A", "A", "C", "D"]

pandas/tests/groupby/test_function.py

@@ -4,6 +4,7 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
 from pandas.errors import UnsupportedFunctionCall
 
 import pandas as pd
@@ -1004,6 +1005,7 @@ def test_frame_describe_unstacked_format():
     tm.assert_frame_equal(result, expected)
 
 
+@pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
 @pytest.mark.filterwarnings(
     "ignore:"
     "indexing past lexsort depth may impact performance:"

pandas/tests/groupby/test_quantile.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 import pandas as pd
 from pandas import DataFrame, Index
 import pandas._testing as tm
@@ -71,6 +73,7 @@ def test_quantile_array():
     tm.assert_frame_equal(result, expected)
 
 
+@pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
 def test_quantile_array2():
     # https://github.com/pandas-dev/pandas/pull/28085#issuecomment-524066959
     df = DataFrame(
@@ -106,6 +109,7 @@ def test_quantile_array_no_sort():
     tm.assert_frame_equal(result, expected)
 
 
+@pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
 def test_quantile_array_multiple_levels():
     df = DataFrame(
         {"A": [0, 1, 2], "B": [3, 4, 5], "c": ["a", "a", "a"], "d": ["a", "a", "b"]}
@@ -216,6 +220,8 @@ def test_quantile_missing_group_values_correct_results(
 @pytest.mark.parametrize("q", [0.5, [0.0, 0.5, 1.0]])
 def test_groupby_quantile_nullable_array(values, q):
     # https://github.com/pandas-dev/pandas/issues/33136
+    if isinstance(q, list):
+        pytest.skip("GH#39089 Numpy changed dtype inference")
     df = DataFrame({"a": ["x"] * 3 + ["y"] * 3, "b": values})
     result = df.groupby("a")["b"].quantile(q)
 
@@ -256,6 +262,7 @@ def test_groupby_timedelta_quantile():
     tm.assert_frame_equal(result, expected)
 
 
+@pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
 def test_columns_groupby_quantile():
     # GH 33795
     df = DataFrame(

pandas/tests/indexes/base_class/test_indexing.py

@@ -1,6 +1,8 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
+
 from pandas import Index
 import pandas._testing as tm
 
@@ -13,6 +15,7 @@ def test_get_slice_bounds_within(self, kind, side, expected):
         result = index.get_slice_bound("e", kind=kind, side=side)
         assert result == expected
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     @pytest.mark.parametrize("kind", ["getitem", "loc", None])
     @pytest.mark.parametrize("side", ["left", "right"])
     @pytest.mark.parametrize(

pandas/tests/indexes/base_class/test_where.py

@@ -1,10 +1,14 @@
 import numpy as np
+import pytest
+
+from pandas.compat import is_numpy_dev
 
 from pandas import Index
 import pandas._testing as tm
 
 
 class TestWhere:
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_where_intlike_str_doesnt_cast_ints(self):
         idx = Index(range(3))
         mask = np.array([True, False, True])

pandas/tests/indexes/categorical/test_indexing.py

@@ -1,6 +1,7 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
 from pandas.errors import InvalidIndexError
 
 import pandas as pd
@@ -129,6 +130,7 @@ def test_take_invalid_kwargs(self):
 
 
 class TestGetLoc:
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_get_loc(self):
         # GH 12531
         cidx1 = CategoricalIndex(list("abcde"), categories=list("edabc"))

pandas/tests/indexes/interval/test_interval.py

@@ -4,6 +4,7 @@
 import numpy as np
 import pytest
 
+from pandas.compat import is_numpy_dev
 from pandas.errors import InvalidIndexError
 
 import pandas as pd
@@ -916,6 +917,9 @@ def test_searchsorted_different_argument_classes(klass):
     tm.assert_numpy_array_equal(result, expected)
 
 
+@pytest.mark.xfail(
+    is_numpy_dev, reason="GH#39089 Numpy changed dtype inference", strict=False
+)
 @pytest.mark.parametrize(
     "arg", [[1, 2], ["a", "b"], [Timestamp("2020-01-01", tz="Europe/London")] * 2]
 )

pandas/tests/indexes/period/test_constructors.py

@@ -2,6 +2,7 @@
 import pytest
 
 from pandas._libs.tslibs.period import IncompatibleFrequency
+from pandas.compat import is_numpy_dev
 
 from pandas.core.dtypes.dtypes import PeriodDtype
 
@@ -304,6 +305,7 @@ def test_constructor_incompat_freq(self):
                 )
             )
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_constructor_mixed(self):
         idx = PeriodIndex(["2011-01", NaT, Period("2011-01", freq="M")])
         exp = PeriodIndex(["2011-01", "NaT", "2011-01"], freq="M")

pandas/tests/indexes/test_base.py

@@ -9,7 +9,7 @@
 import pytest
 
 from pandas._libs.tslib import Timestamp
-from pandas.compat import IS64
+from pandas.compat import IS64, is_numpy_dev
 from pandas.compat.numpy import np_datetime64_compat
 from pandas.util._test_decorators import async_mark
 
@@ -1358,6 +1358,7 @@ def test_slice_float_locs(self, dtype):
         assert index2.slice_locs(8.5, 1.5) == (2, 6)
         assert index2.slice_locs(10.5, -1) == (0, n)
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_slice_locs_dup(self):
         index = Index(["a", "a", "b", "c", "d", "d"])
         assert index.slice_locs("a", "d") == (0, 6)
@@ -1397,6 +1398,9 @@ def test_slice_locs_na_raises(self):
         with pytest.raises(KeyError, match=""):
             index.slice_locs(end=1.5)
 
+    @pytest.mark.xfail(
+        is_numpy_dev, reason="GH#39089 Numpy changed dtype inference", strict=False
+    )
     @pytest.mark.parametrize(
         "in_slice,expected",
         [

pandas/tests/indexes/test_engines.py

@@ -4,6 +4,7 @@
 import pytest
 
 from pandas._libs import algos as libalgos, index as libindex
+from pandas.compat import is_numpy_dev
 
 import pandas as pd
 import pandas._testing as tm
@@ -198,6 +199,7 @@ def test_is_unique(self):
         engine = self.engine_type(lambda: arr, len(arr))
         assert engine.is_unique is False
 
+    @pytest.mark.xfail(is_numpy_dev, reason="GH#39089 Numpy changed dtype inference")
     def test_get_loc(self):
         # unique
         arr = np.array(self.values, dtype=self.dtype)