So I tried it again in standalone mode (spark-shell) and the df.observe() functionality works. I tried sum, count, conditional aggregations using 'when', etc and all of this works in spark-shell. But, with spark-on-k8s, cluster mode, only using lit() as the aggregation column works. No other aggregation, including, count, sum, etc work.
-- Sent from: http://apache-spark-user-list.1001560.n3.nabble.com/ --------------------------------------------------------------------- To unsubscribe e-mail: user-unsubscr...@spark.apache.org