dario-liberman commented on issue #11775: URL: https://github.com/apache/pinot/issues/11775#issuecomment-1762670074
@jackjlli - I see in https://github.com/apache/pinot/issues/7213 you wanted to be able to calculate aggregations on moving time windows, such as avg COVID cases in a 7 day moving window. Here a proposal to solve it with a very simple transform function which works already in single stage engine thanks to multivalued group by support in Pinot. Note also that the traditional SQL window semantics discussed as answer to your use case are actually not very good at sliding time windows, here instead we refer for analogy to streaming SQL constructs which capture timeseries semantics much better. I actually believe moving time window aggregates for distinct counts with traditional SQL can not be calculated with row window aggregation, most answers out there do a left join on itself by time column between time window boundaries, with likely edge cases for time gaps in the data; a surprisingly difficult task. cc: @kishoreg -- 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: commits-unsubscr...@pinot.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
