spark-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From nseggert <nicholas.egg...@target.com>
Subject SPARK-17455 Isotonic Regression fix languishing
Date Tue, 13 Dec 2016 16:31:23 GMT
I have PR that has been sitting untouched for months. Could someone please
take a look at it?

https://github.com/apache/spark/pull/15018 (SPARK-17455) in JIRA.

This PR fixes problems with the way isotonic regression was implemented that
caused it to take exponential time for some inputs. This problem made it
unusable for me, because I occasionally hit those cases. I'd assume it's
very likely to affect others as well. 

My patch just translates the scikit-learn implementation into Scala, so I
think it should be fairly uncontroversial.

Also note that although this is a change to the mllib code, this is one of
those cases where the ml implementation just calls the mllib version.



--
View this message in context: http://apache-spark-developers-list.1001551.n3.nabble.com/SPARK-17455-Isotonic-Regression-fix-languishing-tp20212.html
Sent from the Apache Spark Developers List mailing list archive at Nabble.com.

---------------------------------------------------------------------
To unsubscribe e-mail: dev-unsubscribe@spark.apache.org


Mime
View raw message