Github user mengxr commented on the pull request:
https://github.com/apache/spark/pull/3519#issuecomment-67088697
@zapletal-martin Some high-level comments:
1. The implementation introduces APIs that are not necessary for this PR.
For example, `WeightedLabeledPoint` and `MonotonicityConstraint`. It should be
sufficient to have only `IsotonicRegression` and `IsotonicRegressionModel`.
Please try to make the public APIs minimal.
2. I think it would be better if an `IsotonicRegression` takes an
`RDD[(Double, Double)]` instead of `RDD[WeightedLabeledPoint]` and use the
natural ordering. It is easy for the new pipeline API, where
`IsotonicRegression` expects 2 or 3 (with weight) columns and the model applies
to a single column.
---
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]
