jorisvandenbossche commented on pull request #7891: URL: https://github.com/apache/arrow/pull/7891#issuecomment-670478968
cc @fjetter @xhochy in case if are familiar with those failures There are several failures like this (related to exit codes): ``` _________________________________ test_simple __________________________________ cli = <function cli.<locals>._cli at 0x7ff27b5fed40> built_cube = Cube(dimension_columns=('x', 'y'), partition_columns=('p', 'q'), uuid_prefix='my_cube', seed_dataset='source', index_columns=frozenset({'i2', 'i1'})) mock_prompt = <function mock_prompt.<locals>._f at 0x7ff27b5b6830> df_complete = i1 i2 p ... v2 x y 0 True 2018-01-01 0 ... ... 8, 9] 1 0 3 False 2019-01-01 0 ... [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,... 1 1 [4 rows x 9 columns] skv = None def test_simple(cli, built_cube, mock_prompt, df_complete, skv): mock_prompt(["", ""]) # conditions # payload result = cli("--store=cubes", "my_cube", "query", input="df\n") str_df = str(df_complete.loc[:, ["p", "q", "x", "y"]].reset_index(drop=True)) > assert result.exit_code == 1 E assert 2 == 1 E + where 2 = <Result SystemExit(2)>.exit_code /kartothek/tests/cli/test_query.py:64: AssertionError ``` and there are a few failures like ``` ________ test_predicate_pushdown_null_col[True-serialiser4-df3-value3] _________ store = <conftest.FakeStore object at 0x7ff242feef50>, df = t 0 NaT value = Timestamp('2017-01-01 00:00:00'), predicate_pushdown_to_io = True serialiser = CsvSerializer(compress=True) @pytest.mark.parametrize( "df,value", [ (pd.DataFrame({"u": pd.Series([None], dtype=object)}), "foo"), (pd.DataFrame({"b": pd.Series([None], dtype=object)}), b"foo"), (pd.DataFrame({"f": pd.Series([np.nan], dtype=float)}), 1.2), ( pd.DataFrame({"t": pd.Series([pd.NaT], dtype="datetime64[ns]")}), pd.Timestamp("2017"), ), ], ) @predicate_serialisers @pytest.mark.parametrize("predicate_pushdown_to_io", [True, False]) def test_predicate_pushdown_null_col( store, df, value, predicate_pushdown_to_io, serialiser ): key = serialiser.store(store, "prefix", df) expected = df.iloc[[]].copy() predicates = [[(df.columns[0], "==", value)]] result = serialiser.restore_dataframe( store, key, predicate_pushdown_to_io=predicate_pushdown_to_io, predicates=predicates, ) pdt.assert_frame_equal( result.reset_index(drop=True), expected.reset_index(drop=True), > check_dtype=serialiser.type_stable, ) tests/serialization/test_dataframe.py:529: _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ left = array([], dtype=float64) right = <DatetimeArray> [] Length: 0, dtype: datetime64[ns], check_dtype = True index_values = array([], dtype=int64) check_less_precise = <object object at 0x7ff2986040f0>, check_exact = False rtol = 1e-05, atol = 1e-08 > assert isinstance(left, ExtensionArray), "left is not an ExtensionArray" E AssertionError: left is not an ExtensionArray /opt/conda/envs/arrow/lib/python3.7/site-packages/pandas/_testing.py:1175: AssertionError ``` Since this is related to a pandas testing utility, does kartothek already support pandas 1.1? ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org