Nick Becker
Nick Becker
Hi! I just came across this issue due to the cuML / RAPIDS mention. I wanted to note that we've implemented input-to-output data type consistency for all cuML estimators (not...
Noticed an example in which the small model fails but the medium model succeeds. `murmured` is incorrectly tagged as a proper noun starting the sentence in the example below (perhaps...
Done, thanks for the bump.
I see some test failures in test_parquet.py. Are any of the parquet tests known to be flaky, or should I look into any unexpected interaction with this PR? I also...
Test failures appear tied to the version of pandas. The Python 3.9 environment uses pandas 1.4, in which groupby.rank behaves differently than in prior versions. The 3.8 environment uses 1.2.5...
I've been thinking about this a bit more, and I've arrived a question that I think applies to this PR and also the already implemented `groupby.{shift, transform, apply}` (though is...
While cuDF could raise a more informative error rather than leaking internals, this is a Dask issue due to not being able to align the indexes. We can leave this...
cc @vibhujawa
As a note, [membership_vector and all_points_membership_vectors](https://github.com/rapidsai/cuml/issues/4724) are on our radar for cuML's HDBSCAN. Perhaps this might be an opportunity to define something like `is_hdbscan_like` in the spirit of scikit-learn's `is_classifier`...