[
https://issues.apache.org/jira/browse/SPARK-13944?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15240322#comment-15240322
]
Xiangrui Meng commented on SPARK-13944:
---------------------------------------
There are more production workflows using RDD-based APIs than DataFrame-based
APIs since many users are still running Spark 1.4 or earlier. It would be nice
if we can keep binary compatibility on RDD-based APIs in Spark 2.0. Using type
alias is not a good solution because 1) it is not Java-compatible, 2) it
introduces dependency from the RDD-based API to mllib-local, which means future
development on mllib-local might cause behavior changes or break changes to the
RDD-based API. Since we already decided that the RDD-based API would go into
maintenance mode in Spark 2.0. Leaving some old code there won't increase
maintenance cost, compared with the type alias.
We can provide a converter than converts all `mllib.linalg` types to
`ml.linalg` types in Spark 2.0 to help users migrate to `ml.linalg`.
> Separate out local linear algebra as a standalone module without Spark
> dependency
> ---------------------------------------------------------------------------------
>
> Key: SPARK-13944
> URL: https://issues.apache.org/jira/browse/SPARK-13944
> Project: Spark
> Issue Type: New Feature
> Components: Build, ML
> Affects Versions: 2.0.0
> Reporter: Xiangrui Meng
> Assignee: DB Tsai
> Priority: Blocker
>
> Separate out linear algebra as a standalone module without Spark dependency
> to simplify production deployment. We can call the new module
> spark-mllib-local, which might contain local models in the future.
> The major issue is to remove dependencies on user-defined types.
> The package name will be changed from mllib to ml. For example, Vector will
> be changed from `org.apache.spark.mllib.linalg.Vector` to
> `org.apache.spark.ml.linalg.Vector`. The return vector type in the new ML
> pipeline will be the one in ML package; however, the existing mllib code will
> not be touched. As a result, this will potentially break the API. Also, when
> the vector is loaded from mllib vector by Spark SQL, the vector will
> automatically converted into the one in ml package.
--
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
