[jira] [Commented] (SPARK-9716) BinaryClassificationEvaluator should accept Double prediction column

["Luvsandondov Lkhamsuren (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-9716?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14732119#comment-14732119 ] 

Luvsandondov Lkhamsuren commented on SPARK-9716:
------------------------------------------------

I was thinking on having method that checks if the dataType equals one of the list of dataTypes inside SchemaUtils (similar to checkColumnType), 
{code:title=SchemUtils.scala | borderStyle=solid}
 def checkColumnType(schema: StructType, colName: String, dataTypes: Array[DataType], msg: String = "") = ???
{code}
then use the method to check if the rawPredictionColumn is either Vector or Double as follows:

{code:title=BinaryClassificationEvaluator.scala | borderStyle=solid}
 override def evaluate(dataset: DataFrame): Double = {
    val schema = dataset.schema
    SchemaUtils.checkColumnType(schema, $(rawPredictionCol), Array(new VectorUDT, DoubleType))
    SchemaUtils.checkColumnType(schema, $(labelCol), DoubleType)

    val scoreAndLabels = dataset.select($(rawPredictionCol), $(labelCol))
      .map ( row => row match {
        case Row(rawPrediction: Vector, label: Double) =>
        (rawPrediction(1), label)
        case Row(rawPrediction: Double, label: Double) => 
        (rawPrediction, label)
       }
      )
    val metrics = new BinaryClassificationMetrics(scoreAndLabels)
    val metric = $(metricName) match {
      case "areaUnderROC" => metrics.areaUnderROC()
      case "areaUnderPR" => metrics.areaUnderPR()
    }
    metrics.unpersist()
    metric
  }
{code}

How does it sound?

> BinaryClassificationEvaluator should accept Double prediction column
> --------------------------------------------------------------------
>
>                 Key: SPARK-9716
>                 URL: https://issues.apache.org/jira/browse/SPARK-9716
>             Project: Spark
>          Issue Type: Improvement
>          Components: ML
>            Reporter: Joseph K. Bradley
>            Priority: Minor
>
> BinaryClassificationEvaluator currently expects the rawPrediction column, of type Vector.  It should also accept a Double prediction column, with a different set of supported metrics.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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