Github user sethah commented on the issue:
https://github.com/apache/spark/pull/16699
I'm finding R's behavior for prediction with offsets to be a bit strange.
Yes, R does use the original offsets supplied during training to do prediction,
but what if I want to make predictions on a completely new dataset, with
completely different offsets? R still tries to use the original ones, which
doesn't make sense to me. Then, maybe predicting on "new" data when you've
trained using offsets doesn't make sense? Regardless, here we are allowing
users to pass in a new dataframe to make predictions on, where they can supply
an offset column for the new "test" data - offsets that are not the same as the
training offsets. This is a feature not provided by R, it seems.
---
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
