Github user manishamde commented on the pull request:
https://github.com/apache/spark/pull/2607#issuecomment-58987769
@jkbradley I already have support for a parameter class called
```BoostingStrategy```. I also have support for a shorter argument list when
the user wants to specify only the minimum number of parameters to train.
However, the long argument list is useful when an advanced user wants access to
all the options. Also, in the other scenario, we end up having a long parameter
to list to call the constructor of the parameter class.
Ideally, a user should not have to worry about using algorithm specific
parameters classes (which we mark as Experimental) to access the algorithms. I
can remove some tree-specific options to fit the limit of 10 at the risk of
making it harder for users to set advanced options. I think the limit of 10 is
arbitrary for external facing API especially since Scala and Python support
named named parameters with default values (unfortunately Java does not).
Having said that, we can separate boosting and underlying algo parameters after
standard MLlib API work.
---
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
