public abstract class AbstractEvaluation extends Object implements LeastSquaresProblem.Evaluation
LeastSquaresProblem.Evaluation that is designed for extension. All of the
 methods implemented here use the methods that are left unimplemented.
 
 TODO cache results?| Modifier and Type | Method and Description | 
|---|---|
| double | getCost()Get the cost. | 
| RealMatrix | getCovariances(double threshold)Get the covariance matrix of the optimized parameters. | 
| double | getRMS()Get the normalized cost. | 
| RealVector | getSigma(double covarianceSingularityThreshold)Get an estimate of the standard deviation of the parameters. | 
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, waitgetJacobian, getPoint, getResidualspublic RealMatrix getCovariances(double threshold)
JTJ matrix,
 where J is the Jacobian matrix. The threshold parameter is a
 way for the caller to specify that the result of this computation should be
 considered meaningless, and thus trigger an exception.getCovariances in interface LeastSquaresProblem.Evaluationthreshold - Singularity threshold.public RealVector getSigma(double covarianceSingularityThreshold)
sd(a[i]) ~= sqrt(C[i][i]), where a[i] is the optimized
 value of the i-th parameter, and C is the covariance matrix.getSigma in interface LeastSquaresProblem.EvaluationcovarianceSingularityThreshold - Singularity threshold (see computeCovariances).public double getRMS()
getRMS in interface LeastSquaresProblem.Evaluationpublic double getCost()
getCost in interface LeastSquaresProblem.EvaluationLeastSquaresProblem.Evaluation.getResiduals()Copyright © 2003–2016 The Apache Software Foundation. All rights reserved.