Details
-
Improvement
-
Status: Resolved
-
Minor
-
Resolution: Fixed
-
1.6.1
-
None
Description
Currently, dedupWithCombiner only support a single column as its groupBy / orderBy keys. you can still leverage multiple keys, but you will need to add them as column with some expression to unify the columns.
I would suggest adding support of multiple columns in the function itself without having to modify the df prior to that