Uploaded image for project: 'Spark'
  1. Spark
  2. SPARK-8598

Implementation of 1-sample, two-sided, Kolmogorov Smirnov Test for RDDs

    XMLWordPrintableJSON

    Details

    • Type: New Feature
    • Status: Resolved
    • Priority: Minor
    • Resolution: Fixed
    • Affects Version/s: None
    • Fix Version/s: 1.5.0
    • Component/s: MLlib
    • Labels:
      None
    • Target Version/s:

      Description

      We have implemented a 1-sample, two-sided version of the Kolmogorov Smirnov test, which tests the null hypothesis that the sample comes from a given continuous distribution. We provide various functions to access the functionality: namely, a function that takes an RDD[Double] of the data and a lambda to calculate the CDF, a function that takes an RDD[Double] and an Iterator[(Double,Double,Double)] => Iterator[Double] which uses mapPartition to provide an optimized way to perform the calculation when the CDF calculation requires a non-serializable object (e.g. the apache math commons real distributions), and finally a function that takes an RDD[Double] and a String name of the theoretical distribution to be used. The appropriate result class has been added, as well as tests to the HypothesisTestSuite

        Attachments

          Issue Links

            Activity

              People

              • Assignee:
                josepablocam Jose Cambronero
                Reporter:
                josepablocam Jose Cambronero
                Shepherd:
                Xiangrui Meng
              • Votes:
                0 Vote for this issue
                Watchers:
                5 Start watching this issue

                Dates

                • Created:
                  Updated:
                  Resolved: