The following simple (pyspark) code fails in Spark 2.0.1:
It only fails with all three arguments to .agg, removing any of them prevents the failure. Similar code in Java also fails in the same way, so it isn't specific to the Python API. It runs without error in Spark 2.0.0, so I suspect it might be caused by the fix to SPARK-16648. I've opened a Jira ticket (SPARK-18172) but thought I'd also post to the mailing list to make sure it's noticed. The full stack trace is below: -- View this message in context: http://apache-spark-developers-list.1001551.n3.nabble.com/AnalysisException-in-first-last-during-aggregation-since-2-0-1-tp19724.html Sent from the Apache Spark Developers List mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe e-mail: dev-unsubscr...@spark.apache.org
