andygrove commented on PR #4254: URL: https://github.com/apache/datafusion-comet/pull/4254#issuecomment-4393406537
> > > DataFusion's grouped aggregate operator has a fast path that uses > > > GroupsAccumulator instead of one Accumulator per group > > > > > > Should we try to wire DF aggregation instead of Comet implementation? or there are any Spark inconsistencies? > > They are not compatible the last time I checked - signed vs unsigned types, IIRC would be good to upstream these to datafusion-spark and then see if they could share common code with the core impl -- 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]
