(self)
| 851 | "which causes the `y` column to have a different type depending on whether pyarrow is installed" |
| 852 | ) |
| 853 | def test_to_dask_dataframe_2D(self): |
| 854 | # Test if 2-D dataset is supplied |
| 855 | w = np.random.randn(2, 3) |
| 856 | ds = Dataset({"w": (("x", "y"), da.from_array(w, chunks=(1, 2)))}) |
| 857 | ds["x"] = ("x", np.array([0, 1], np.int64)) |
| 858 | ds["y"] = ("y", list("abc")) |
| 859 | |
| 860 | # dask dataframes do not (yet) support multiindex, |
| 861 | # but when it does, this would be the expected index: |
| 862 | exp_index = pd.MultiIndex.from_arrays( |
| 863 | [[0, 0, 0, 1, 1, 1], ["a", "b", "c", "a", "b", "c"]], names=["x", "y"] |
| 864 | ) |
| 865 | expected = pd.DataFrame({"w": w.reshape(-1)}, index=exp_index) |
| 866 | # so for now, reset the index |
| 867 | expected = expected.reset_index(drop=False) |
| 868 | actual = ds.to_dask_dataframe(set_index=False) |
| 869 | |
| 870 | assert isinstance(actual, dd.DataFrame) |
| 871 | assert_frame_equal(actual.compute(), expected) |
| 872 | |
| 873 | @pytest.mark.xfail(raises=NotImplementedError) |
| 874 | def test_to_dask_dataframe_2D_set_index(self): |
nothing calls this directly
no test coverage detected