Github user MLnick commented on the issue:
https://github.com/apache/spark/pull/17384
For the other algos we've done something similar - but aimed to allowing
any numeric type for `label` and `weight` columns. I'm not totally against this
here, but I think it falls into a slightly different "category" - i.e.
supporting different numeric types for input features.
Overall I think it's more natural to enforce a double feature column (if
you think about it, a feature vector column is implicitly enforced to be double
as we don't support arbitrary numeric types for vectors).
I'm not convinced this solves any major pain point.
---
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]
