ENH Array API support for euclidean_distances and rbf_kernel#29433
ENH Array API support for euclidean_distances and rbf_kernel#29433ogrisel merged 30 commits intoscikit-learn:mainfrom
Conversation
|
@ogrisel Don't we have the latest version of array api strict on all CI pipelines? This seems to be failing for some pipelines probably because of the older version of array-api-strict, while it works fine on my local system with the latest version of array-api-strict. Should we fix this just for supporting an old version? |
|
@OmarManzoor let me update this PR with Further CI fixes will follow-up once #29436 is reviewed and merged. |
|
@ogrisel These CI errors seem a bit weird. One is related to the docstring of |
|
I thought I fixed that docstring failure somewhere else already. EDIT: here it is: @OmarManzoor feel free to steal this from the Ridge PR since I won't have time to finalize it this week. |
|
Please feel free to merge |
ogrisel
left a comment
There was a problem hiding this comment.
I labeled "CUDA CI" to trigger a CI run with a CUDA GPU.
If everything is green, LGTM.
|
I tried to relabel this PR to trigger the CUDA CI out of curiosity but it's still waiting on workers for some unknown reason: |
adrinjalali
left a comment
There was a problem hiding this comment.
Otherwise LGTM. Thanks @ogrisel for the ping.
adrinjalali
left a comment
There was a problem hiding this comment.
LGTM. I'll let @ogrisel have another look and merge.
|
@ogrisel Can this be merged? |
Reference Issues/PRs
Towards #26024
What does this implement/fix? Explain your changes.
Any other comments?