Github user srowen commented on the issue:
https://github.com/apache/spark/pull/20164
OK, the code in question uses the `rawPredictionCol` from the `models` it
is given. Yes they'd have to be unique for this to make sense, because it adds
those raw prediction columns to the output. Isn't this always true though? You
can't have a pipeline where elements add a column that exist, or were
previously added by the pipeline.
I'm missing why it's not a problem to alter the output's raw prediction
column from what the user had specified in `models`?
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
