thisisnic commented on issue #45139: URL: https://github.com/apache/arrow/issues/45139#issuecomment-7039294421
This still reproduces on 26.0.0: with the reprex above, every sample of 10 from `1:5000` comes back sorted and below about 280, and the two-row reprex gives roughly a 70/30 split with only around 800 rows from 1000 draws. It's the same underlying problem as #38638 (asking for `n` rows oversamples by proportion and then takes the first `n` rows that pass, which favours the start of the table and can return too few rows), so I'm closing this as a duplicate of that issue. A fix is in progress in #51385. -- 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]
