Package dk.alexandra.fresco.stat.tests

Class KruskallWallisTest

Object

KruskallWallisTest

All Implemented Interfaces:

Computation<SFixed,ProtocolBuilderNumeric>

public class KruskallWallisTest
extends Object
implements Computation<SFixed,ProtocolBuilderNumeric>

Compute the Kruskall-Wallis test statistic on k groups, also known as one-way ANOVA on ranks. The test statistic may be approximated by a 𝛘² distribution with k-1 degrees of freedom.

The elements are ranked sequentially ignoring ties unless the averageTies parameter is set, in which case ties are replaced with the average of the ranks of the elements it is tied with. Setting averageTies to true returns the standard test statistics, but using false should not affect the test statistic too much. If averageTies is set to true, the ranks of the ties will be leaked to all parties.

Constructor Summary

Constructors

Constructor	Description
KruskallWallisTest(List<List<DRes<SInt>>> observed)

Method Summary

Modifier and Type	Method	Description
DRes<SFixed>	buildComputation(ProtocolBuilderNumeric builder)
static List<List<DRes<SInt>>>	fromSFixed(List<List<DRes<SFixed>>> observed)	If the test is to be applied on fixed point numbers (SFixed's), this method should be used to transform the data,

Methods inherited from class Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

KruskallWallisTest

public KruskallWallisTest(List<List<DRes<SInt>>> observed)

Method Details

fromSFixed

public static List<List<DRes<SInt>>> fromSFixed(List<List<DRes<SFixed>>> observed)

If the test is to be applied on fixed point numbers (SFixed's), this method should be used to transform the data,

Parameters:

observed - The data as fixed point numbers.

Returns:

The input data scaled to integers to be used in the test.

buildComputation

public DRes<SFixed> buildComputation(ProtocolBuilderNumeric builder)

Specified by:

buildComputation in interface Computation<SFixed,ProtocolBuilderNumeric>