pitrou edited a comment on pull request #10520: URL: https://github.com/apache/arrow/pull/10520#issuecomment-861504764
I'm not really surprised and was actually expecting worse results :-) There are two factors: * the reshuffling (row-oriented vs. column-oriented) is an intrinsic problem of this operation and will stress the memory subsystem much more (while `binary_join` is purely sequential) * the implementation iterates row a time, and therefore does dynamic switching between various cases (scalar, array, null options...) in each loop iteration -- 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. For queries about this service, please contact Infrastructure at: [email protected]
