pepijnve commented on issue #11570: URL: https://github.com/apache/datafusion/issues/11570#issuecomment-3860506484
In the context of the review of #19994 it became clear that some further optimisation work could be done in the general 'expression or expression' code path. That has been merged in #20097. What can still be improved further, and would benefit divide by zero protection with few zero denominators, is the `scatter` function. Profiling shows that in the "10% zeroes" benchmark, that `scatter` chews up just over 50% of elapsed time. <img width="1410" height="251" alt="Image" src="https://github.com/user-attachments/assets/06ec6802-ecc9-449a-a1b9-843b7176aab2" /> -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
