sunchao commented on issue #7095: URL: https://github.com/apache/datafusion/issues/7095#issuecomment-2305877287
@Lordworms @alamb the code is still in my branch: https://github.com/sunchao/arrow-datafusion/commits/vectorized-hash-table/ and hasn't been updated for a while, so not sure how easy it can rebase. One missing piece for the work is to eliminate the `convert_columns` and `convert_rows` cost. Last time I measured it took ~30% of the total time. For this, I think we need to extend the Array [builder](https://github.com/apache/arrow-rs/tree/master/arrow-array/src/builder) to add a get API too, so that while we are building the group values (which is represented via Arrow vectors) we can still look into the elements in various operations such as equality check. -- 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]
