pitrou commented on pull request #11019: URL: https://github.com/apache/arrow/pull/11019#issuecomment-914071537
> Some conclusions: Both implementations (partition_nth:intro_select and partition_nth:heap) show similar numbers. Interesting. Did you double-check that no mistake was made when running the benchmarks (i.e. are you really comparing the two different algorithms)? > IMO, implementation of stable algorithms need more exploration, and it could be implemented in a separate follow up JIRA issue Hmm, why not, but `sort_indices` and `partition_nth_indices` are guaranteed to be stable, so it would be better if topK/bottomK wasn't inconsistent. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
