Github user yanboliang commented on the issue:
https://github.com/apache/spark/pull/18613
@felixcheung We don't silently drop features, we use ```handleInvalid``` to
let users decide how to handle invalid features or label. The behavior is
consistent with Scala which supports to handle invalid features and label. BTW,
this PR add support to handle invalid values for label, the original PR by
@wangmiao1981 has already supported to handle invalid features.
With regards to different policy, this may should be discussed. We can
provide two options for features and label respectively, but I think handling
invalid values for features and label with the same policy may be the most
common use case, and I don't want to make code much complicated before we can
see users' strong requirements. And we can split it into two params to control
features and label later when it's really necessary. However, I don't have
strong opinions about this, and looking forward to hear your thoughts. 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]
