Github user jkbradley commented on the issue:
https://github.com/apache/spark/pull/15788
@felixcheung @yanboliang + @shivaram @mengxr Just saw this change. IIRC
we intentionally made spark.glm match R's glm behavior for the binomial family,
though I see it was a little different: glm outputs log-space predictions
(rawPrediction) by default, and spark.glm output probabilities by default
before this PR. This PR makes spark.glm match Scala/Java/Python MLlib APIs
instead of R libraries.
We should pick a clear strategy here: match the rest of MLlib or match
familiar R libraries.
---
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 [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
