org.apache.commons.math.optimization.general
Class GaussNewtonOptimizer

java.lang.Object
  org.apache.commons.math.optimization.direct.BaseAbstractVectorialOptimizer<DifferentiableMultivariateVectorialFunction>
      org.apache.commons.math.optimization.general.AbstractLeastSquaresOptimizer
          org.apache.commons.math.optimization.general.GaussNewtonOptimizer

All Implemented Interfaces:

BaseMultivariateVectorialOptimizer<DifferentiableMultivariateVectorialFunction>, BaseOptimizer<VectorialPointValuePair>, DifferentiableMultivariateVectorialOptimizer

public class GaussNewtonOptimizer
extends AbstractLeastSquaresOptimizer

Gauss-Newton least-squares solver.

This class solve a least-square problem by solving the normal equations of the linearized problem at each iteration. Either LU decomposition or QR decomposition can be used to solve the normal equations. LU decomposition is faster but QR decomposition is more robust for difficult problems.

Since:

2.0

Version:

$Id: GaussNewtonOptimizer.java 1175100 2011-09-24 04:47:38Z celestin $

Field Summary

Fields inherited from class org.apache.commons.math.optimization.general.AbstractLeastSquaresOptimizer

cols, cost, objective, point, residuals, rows, weightedResidualJacobian, weightedResiduals

Fields inherited from class org.apache.commons.math.optimization.direct.BaseAbstractVectorialOptimizer

evaluations

Constructor Summary

GaussNewtonOptimizer()
Simple constructor with default settings.

GaussNewtonOptimizer(boolean useLU)
Simple constructor with default settings.

GaussNewtonOptimizer(boolean useLU, ConvergenceChecker<VectorialPointValuePair> checker)

GaussNewtonOptimizer(ConvergenceChecker<VectorialPointValuePair> checker)
Simple constructor with default settings.

Method Summary

VectorialPointValuePair

doOptimize() Perform the bulk of the optimization algorithm.

Methods inherited from class org.apache.commons.math.optimization.general.AbstractLeastSquaresOptimizer

getChiSquare, getCovariances, getCovariances, getJacobianEvaluations, getRMS, guessParametersErrors, optimize, updateJacobian, updateResidualsAndCost

Methods inherited from class org.apache.commons.math.optimization.direct.BaseAbstractVectorialOptimizer

computeObjectiveValue, getConvergenceChecker, getEvaluations, getMaxEvaluations, getStartPoint, getTargetRef, getWeightRef, setConvergenceChecker

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.commons.math.optimization.BaseOptimizer

getConvergenceChecker, getEvaluations, getMaxEvaluations, setConvergenceChecker

Constructor Detail

GaussNewtonOptimizer

public GaussNewtonOptimizer()

Simple constructor with default settings. The normal equations will be solved using LU decomposition and the convergence check is set to a SimpleVectorialValueChecker with default tolerances.

GaussNewtonOptimizer

public GaussNewtonOptimizer(ConvergenceChecker<VectorialPointValuePair> checker)

Simple constructor with default settings. The normal equations will be solved using LU decomposition.

Parameters:

checker - Convergence checker.

GaussNewtonOptimizer

public GaussNewtonOptimizer(boolean useLU)

Simple constructor with default settings. The convergence check is set to a SimpleVectorialValueChecker with default tolerances.

Parameters:

useLU - If true, the normal equations will be solved using LU decomposition, otherwise they will be solved using QR decomposition.

GaussNewtonOptimizer

public GaussNewtonOptimizer(boolean useLU,
                            ConvergenceChecker<VectorialPointValuePair> checker)

Parameters:

useLU - If true, the normal equations will be solved using LU decomposition, otherwise they will be solved using QR decomposition.

checker - Convergence checker.

Method Detail

doOptimize

public VectorialPointValuePair doOptimize()

Perform the bulk of the optimization algorithm.

Specified by:

doOptimize in class BaseAbstractVectorialOptimizer<DifferentiableMultivariateVectorialFunction>

Returns:

the point/value pair giving the optimal value for the objective function.