srowen commented on a change in pull request #26905: [SPARK-30266][SQL] Avoid
overflow and match error in ApproximatePercentile
URL: https://github.com/apache/spark/pull/26905#discussion_r358578200
##########
File path:
sql/core/src/test/scala/org/apache/spark/sql/ApproximatePercentileQuerySuite.scala
##########
@@ -182,7 +182,7 @@ class ApproximatePercentileQuerySuite extends QueryTest
with SharedSparkSession
spark.sql(
s"""SELECT
| key,
- | percentile_approx(null, 0.5)
+ | percentile_approx(cast(null as int), 0.5)
Review comment:
It doesn't make much sense, yeah. I am not sure whether we generally return
null or an error in a case like this. Following Hive seems like a reasonable
guide. It could be OK to change behavior if there's a release note attached to
the JIRA, and if we think it's more bug fix than anything
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
