Github user jkbradley commented on the pull request:
https://github.com/apache/spark/pull/5980#issuecomment-99977826
One issue was brought up: What about ranges of buckets where the ends
(-inf, firstSplit] and (lastSplit, inf) don't make sense? E.g., for ages, we
don't want a range (-inf, 0]. We could include this in a few ways:
* Make users include -inf and/or inf in the splits if they want those
ranges.
* Have Boolean parameters lowerInclusive, upperInclusive to indicate
whether to include those ranges.
I prefer the Boolean parameters with defaults of true. That will make it
easy to have the default behavior accept all numerical values. When the range
does not to to -inf and inf, then we'll have to throw an error and report the
bad value.
Also, Bucketizer should extend UnaryTransformer.
Thanks!
---
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]
