Github user maryannxue commented on the issue:
https://github.com/apache/spark/pull/21699
Thank you for the change, @MaxGekk!
@HyukjinKwon my idea was actually that the overloaded versions of pivot
would be `pivot(column: Column, values, Seq[Column])`, so that we can construct
different types in "values". The constant check will be done in Analyzer, so we
don't need to worry about it here.
Ultimately we would like to support complex-typed values in `pivot(column:
Column)` as well, but I think we can make this in a different PR.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
