[jira] [Updated] (SPARK-29815) Missing persist in ml.tuning.CrossValidator.fit()

["Aman Omer (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/SPARK-29815?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Aman Omer updated SPARK-29815:
------------------------------
        Parent: SPARK-29818
    Issue Type: Sub-task  (was: Improvement)

> Missing persist in ml.tuning.CrossValidator.fit()
> -------------------------------------------------
>
>                 Key: SPARK-29815
>                 URL: https://issues.apache.org/jira/browse/SPARK-29815
>             Project: Spark
>          Issue Type: Sub-task
>          Components: ML
>    Affects Versions: 2.4.3
>            Reporter: Dong Wang
>            Priority: Major
>
> dataset.toDF.rdd in ml.tuning.CrossValidator.fit(dataset: Dataset[_]) will generate two rdds: training and validation. Some actions will be operated on these two rdds, but dataset.toDF.rdd is not persisted, which will cause recomputation.
> {code:scala}
>     // Compute metrics for each model over each split
>     val splits = MLUtils.kFold(dataset.toDF.rdd, $(numFolds), $(seed)) // dataset.toDF.rdd should be persisted
>     val metrics = splits.zipWithIndex.map { case ((training, validation), splitIndex) =>
>       val trainingDataset = sparkSession.createDataFrame(training, schema).cache()
>       val validationDataset = sparkSession.createDataFrame(validation, schema).cache()
> {scala}
> This issue is reported by our tool CacheCheck, which is used to dynamically detecting persist()/unpersist() api misuses.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

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