SuanShu, a Java numerical and statistical library

Class Glejser

  extended by com.numericalmethod.suanshu.stats.test.HypothesisTest
      extended by com.numericalmethod.suanshu.stats.test.regression.linear.heteroskedasticity.Heteroskedasticity
          extended by com.numericalmethod.suanshu.stats.test.regression.linear.heteroskedasticity.Glejser

public class Glejser
extends Heteroskedasticity

The Glejser test tests for conditional heteroskedasticity. The test statistics is computed by regressing absolute residuals from the original regression against the original regressors (plus intercept). The test is a chi-squared test: the test statistic distribution is nχ2 with k degrees of freedom.

Field Summary
Fields inherited from class com.numericalmethod.suanshu.stats.test.regression.linear.heteroskedasticity.Heteroskedasticity
N, residuals
Constructor Summary
Glejser(LMResiduals residuals)
          Perform the Glejser test to test for heteroskedasticity in a linear regression model.
Method Summary
 OLSRegression getAuxiliaryRegression()
          Define the transformation of residuals.
Methods inherited from class com.numericalmethod.suanshu.stats.test.regression.linear.heteroskedasticity.Heteroskedasticity
getAlternativeHypothesis, getAuxiliaryOLSRegression, getNullHypothesis, getX2, pValue, statistics
Methods inherited from class com.numericalmethod.suanshu.stats.test.HypothesisTest
isNullRejected, nGroups, nObs, oneSidedPvalue
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail


public Glejser(LMResiduals residuals)
Perform the Glejser test to test for heteroskedasticity in a linear regression model.

residuals - the Residuals object from an OLS regression
Method Detail


public OLSRegression getAuxiliaryRegression()
Description copied from class: Heteroskedasticity
Define the transformation of residuals.

Specified by:
getAuxiliaryRegression in class Heteroskedasticity
an auxiliary regression

Copyright © 2010-2014 Numerical Method Incorporation Limited. All Rights Reserved.