Package org.apache.commons.math3.fitting.leastsquares

Class AbstractEvaluation

java.lang.Object

org.apache.commons.math3.fitting.leastsquares.AbstractEvaluation

All Implemented Interfaces:

LeastSquaresProblem.Evaluation

Direct Known Subclasses:

DenseWeightedEvaluation, LeastSquaresFactory.LocalLeastSquaresProblem.LazyUnweightedEvaluation, LeastSquaresFactory.LocalLeastSquaresProblem.UnweightedEvaluation

public abstract class AbstractEvaluation
extends java.lang.Object
implements LeastSquaresProblem.Evaluation

An implementation of LeastSquaresProblem.Evaluation that is designed for extension. All of the methods implemented here use the methods that are left unimplemented.

TODO cache results?

Since:

3.3

Field Summary

Fields

Modifier and Type	Field	Description
private int	observationSize	number of observations

Constructor Summary

Constructors

Constructor	Description
AbstractEvaluation(int observationSize)	Constructor.

Method Summary

Modifier and Type	Method	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.

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.math3.fitting.leastsquares.LeastSquaresProblem.Evaluation

getJacobian, getPoint, getResiduals

Field Detail

observationSize

private final int observationSize

number of observations

Constructor Detail

AbstractEvaluation

AbstractEvaluation(int observationSize)

Constructor.

Parameters:

observationSize - the number of observation. Needed for getRMS().

Method Detail

getCovariances

public RealMatrix getCovariances(double threshold)

Get the covariance matrix of the optimized parameters.
Note that this operation involves the inversion of the 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.

Specified by:

getCovariances in interface LeastSquaresProblem.Evaluation

Parameters:

threshold - Singularity threshold.

Returns:

the covariance matrix.

getSigma

public RealVector getSigma(double covarianceSingularityThreshold)

Get an estimate of the standard deviation of the parameters. The returned values are the square root of the diagonal coefficients of the covariance matrix, 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.

Specified by:

getSigma in interface LeastSquaresProblem.Evaluation

Parameters:

covarianceSingularityThreshold - Singularity threshold (see computeCovariances).

Returns:

an estimate of the standard deviation of the optimized parameters

getRMS

public double getRMS()

Get the normalized cost. It is the square-root of the sum of squared of the residuals, divided by the number of measurements.

Specified by:

getRMS in interface LeastSquaresProblem.Evaluation

Returns:

the cost.

getCost

public double getCost()

Get the cost.

Specified by:

getCost in interface LeastSquaresProblem.Evaluation

Returns:

the cost.

See Also:

LeastSquaresProblem.Evaluation.getResiduals()