Github user aokolnychyi commented on the issue: https://github.com/apache/spark/pull/19193 @hvanhovell @cloud-fan I think it would be safer to be consistent with other databases and what Spark does for nested aggregate functions. It is really simple to write a subquery to work around any problems. As it is right now, Scenario 2 mentioned above gives a wrong result and Scenario 4 should be double-checked if we want to support nested window functions. Let me known your opinion.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org