Github user zero323 commented on the issue:
https://github.com/apache/spark/pull/17170
My point is that whenever we use `HasFeaturesCol` it is expected that
`featuresCol` is `ml.linalg.VectorUDT`. This creates nice and consistent API.
If we use it for `FPGrowth` it is no longer so clear. Once / if
`PrefixSpan` is implemented we'll get just another input format using the same
`trait` / mixin. If you look at SO you'll see that some users are already
confused what is the expected input for ML algorithms and I believe this can
make it even worse. Just saying...
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org
For additional commands, e-mail: reviews-h...@spark.apache.org
