You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "yuhao yang (JIRA)" <ji...@apache.org> on 2017/03/22 23:34:41 UTC

[jira] [Created] (SPARK-20062) Inconsistent checking on ML estimator/model copy in the unit tests.

yuhao yang created SPARK-20062:
----------------------------------

             Summary: Inconsistent checking on ML estimator/model copy in the unit tests.
                 Key: SPARK-20062
                 URL: https://issues.apache.org/jira/browse/SPARK-20062
             Project: Spark
          Issue Type: Test
          Components: ML
    Affects Versions: 2.1.0
            Reporter: yuhao yang
            Priority: Minor


Currently {code}     
// copied model must have the same parent.
    MLTestingUtils.checkCopy(model)
 {code}
is missing from many unit tests in ml (only 6 appearances in ml.feature). And even for those with the check, we don't have a consistent place for them and they got scattered among different unit tests. Perhaps that's the reason that the check is often forgotten for new features.

Possible options:
1. put it together with save/load
2. put it in the default parameter check (but not all algorithms has this check).






--
This message was sent by Atlassian JIRA
(v6.3.15#6346)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org