[
https://issues.apache.org/jira/browse/SPARK-13030?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15431319#comment-15431319
]
Nick Pentreath edited comment on SPARK-13030 at 8/22/16 6:08 PM:
-----------------------------------------------------------------
Yes I also agree OHE needs to be an {{Estimator}} in order to actually be
usable in a pipeline. Alternative is to have a "stateful" transformer - but IMO
estimator makes more sense here.
The issue we face is that OHE in 2.0 is locked down and we can't break things
now - since it's no longer {{Experimental}}.
Though in many senses this can be viewed as a bug?!
was (Author: mlnick):
Yes I also agree OHE needs to be an {{Estimator}} in order to actually be
usable in a pipeline. Alternative is to have a "stateful" transformer - but IMO
estimator makes more sense here.
The issue we face is that OHE in 2.0 is locked down and we can't break things
now - since it's no longer {{Experimental}}.
> Change OneHotEncoder to Estimator
> ---------------------------------
>
> Key: SPARK-13030
> URL: https://issues.apache.org/jira/browse/SPARK-13030
> Project: Spark
> Issue Type: Improvement
> Components: ML
> Affects Versions: 1.6.0
> Reporter: Wojciech Jurczyk
>
> OneHotEncoder should be an Estimator, just like in scikit-learn
> (http://scikit-learn.org/stable/modules/generated/sklearn.preprocessing.OneHotEncoder.html).
> In its current form, it is impossible to use when number of categories is
> different between training dataset and test dataset.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
