Dandandan edited a comment on issue #790: URL: https://github.com/apache/arrow-datafusion/issues/790#issuecomment-894177971
Yeah Q1 from TCP-H is comparable to the Q1 of the db-benchmark: very little unique keys. The performance improvement listed above (db-benchmark q10 and others) is for a lot of groups which have some other hot paths. I think this benchmark only uses those types you mentioned, so I don't expect anything from that. There are a few other queries in TCP-H that are a bit more challenging on the aggregation side (q3 maybe?), but the benchmarks are all relatively easy as a lot of the data is filtered out on read and aggregate on big groups. -- 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]
