org.codehaus.jet.regression.estimators
Class AkaikeInformationCriterionEstimator

java.lang.Object
  extended by org.codehaus.jet.regression.estimators.AbstractInformationCriterionEstimator
      extended by org.codehaus.jet.regression.estimators.AkaikeInformationCriterionEstimator
All Implemented Interfaces:
InformationCriterionEstimator

public class AkaikeInformationCriterionEstimator
extends AbstractInformationCriterionEstimator

Estimator for the Akaike Information Criterion (AIC)

 AIC(p)= -2T[ln(sigma^2(p)]+2p
 

Author:
Mauro Talevi
See Also:
InformationCriterionEstimator

Constructor Summary
AkaikeInformationCriterionEstimator()
          Creates an AkaikeInformationCriterionEstimator with default regression estimator
AkaikeInformationCriterionEstimator(MultipleLinearRegressionEstimator regression)
          Creates an AkaikeInformationCriterionEstimator with given regression estimator
 
Method Summary
protected  double calculateIC(int p, int T, double var)
          Calculate AIC
 
Methods inherited from class org.codehaus.jet.regression.estimators.AbstractInformationCriterionEstimator
addData, calculateYVariance, createDefaultRegressionEstimator, estimateIC, getSampleSize, minimiseIC, toRegressands, toRegressors
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

AkaikeInformationCriterionEstimator

public AkaikeInformationCriterionEstimator()
Creates an AkaikeInformationCriterionEstimator with default regression estimator


AkaikeInformationCriterionEstimator

public AkaikeInformationCriterionEstimator(MultipleLinearRegressionEstimator regression)
Creates an AkaikeInformationCriterionEstimator with given regression estimator

Parameters:
regression - the GeneralLinearRegression
Method Detail

calculateIC

protected double calculateIC(int p,
                             int T,
                             double var)
Calculate AIC
 AIC(p)= -2T[ln(sigma^2(p)]+2p
 

Specified by:
calculateIC in class AbstractInformationCriterionEstimator
Parameters:
p - the lag order
T - the sample size
var - the sample variance
Returns:
The AIC value


Copyright © 2006-2007. All Rights Reserved.