Raise in P2P if column dtype is wrong
#8167
Conversation
| def test_set_index_p2p_with_existing_index(): | ||
| df = pd.DataFrame({"a": np.random.randint(0, 3, 20)}, index=np.random.random(20)) | ||
| ddf = dd.from_pandas( | ||
| df, | ||
| npartitions=4, | ||
| ) | ||
| with Client() as c: | ||
| with pytest.raises(TypeError, match="_partitions.*integer"): | ||
| ddf.set_index("a", shuffle="p2p") | ||
|
|
||
|
|
||
| def test_sort_values_p2p_with_existing_divisions(): | ||
| "Regression test for #8165" | ||
| df = pd.DataFrame( | ||
| {"a": np.random.randint(0, 3, 20), "b": np.random.randint(0, 3, 20)} | ||
| ) | ||
| ddf = dd.from_pandas( | ||
| df, | ||
| npartitions=4, | ||
| ) | ||
| with Client() as c: | ||
| with dask.config.set({"dataframe.shuffle.method": "p2p"}): | ||
| with pytest.raises(TypeError, match="_partitions.*integer"): | ||
| ddf = ddf.set_index("a").sort_values("b") |
There was a problem hiding this comment.
These two tests need to be updated once dask/dask#10493 is merged.
Unit Test ResultsSee test report for an extended history of previous test failures. This is useful for diagnosing flaky tests. 21 files ± 0 21 suites ±0 10h 40m 28s ⏱️ - 7m 53s For more details on these failures, see this check. Results for commit 6d3a76a. ± Comparison against base commit 20def28. ♻️ This comment has been updated with latest results. |
| from dask.dataframe.core import new_dd_object | ||
|
|
||
| meta = df._meta | ||
| if not pd.api.types.is_integer_dtype(meta[column]): | ||
| raise TypeError( |
There was a problem hiding this comment.
This used to be tested before dask/dask#10493 has been merged.
| with LocalCluster( | ||
| n_workers=2, dashboard_address=":0", loop=loop | ||
| ) as cluster, Client(cluster) as c: | ||
| ddf.set_index("a", shuffle="p2p") |
There was a problem hiding this comment.
Why are you not using the result of this op?
There was a problem hiding this comment.
Whoopsie, changed this test too many times.
Co-authored-by: Patrick Hoefler <61934744+phofl@users.noreply.github.com>
pre-commit run --all-files