Scott Sievert

Thanks, I've merged master. I made the `test_incremental.py#test_basic` weaker by adding a `pytest.xfail` for TimeoutErrors. The tests pass on my own machine w/o timeouts. They sometimes fail on Travis CI;...

It's hanging for me too, not deterministically: `test_basic` completed in about a second in the 2 for 3 runs I've done this morning. Let me do some more debugging.

I've added this diff to `test_basic` to make the test pass locally: ``` diff - while c.futures or s.tasks: # Make sure cleans up cleanly after running - await asyncio.sleep(0.1)...

Hm... I've pushed that change, which involved this diff: ``` diff - while c.futures or s.tasks: + while c.futures: ``` Is that appropriate? [test_client.py#L400-L435][1] seems to be testing how well...

> The docs have contributing guidelines. https://ml.dask.org/contributing.html

Hm.... looks like the error has to do with timeout on windows. I'm not seeing how that's induced by this PR.

Dask-ML v1.7.0 is from Sep. 2020. Try upgrading, I think it's been fixed since then (it certainly raises an error on the `main` branch): https://github.com/dask/dask-ml/blob/67d28b15dfff7869e9a04def203aa129a3540b27/dask_ml/_partial.py#L96-L98

I think it'd be possible to use [Dask-ML's joblib backend][d] because the implementation of `permutation_test_score` relies on joblib: [github.com/scikit-learn/scikit-learn/.../sklearn/model_selection/_validation.py#L1308][1] [1]:https://github.com/scikit-learn/scikit-learn/blob/baf0ea25d6dd034403370fea552b21a6776bef18/sklearn/model_selection/_validation.py#L1308 [d]:https://ml.dask.org/joblib.html

From the docs of [Dask-ML's joblib backend][1]: > Dask can scale these Joblib-backed algorithms out to a cluster of machines by providing an alternative Joblib backend. The following video demonstrates...

Thanks for this use case. I've got some of the related fixes in https://github.com/dask/dask-ml/pull/671 (which will remain a draft until this PR is merged). Please comment in that PR with...