Github user kiszk commented on the issue: https://github.com/apache/spark/pull/13909 I understand what you want to do. I agree with this in this case. This is because the child of `Project` is an array creation. ```` == Physical Plan == *Project [array((value#2 + 1.1), (value#2 + 2.2)) AS array((value + 1.1), (value + 2.2))#10] +- *SerializeFromObject [input[0, double, false] AS value#2] +- Scan ExternalRDDScan[obj#1] ```` I still have two questions. 1. Is this optimization general? If we have additional operations between `Project` and an array creation, what physical plan is generated? 2. How can we know this case at an array creation? An evaluation order of a physical plan tree is bottom-up. Let me clarify these questions.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org