com.mockturtlesolutions.snifflib.invprobs

Class NMSimplex

java.lang.Object

com.mockturtlesolutions.snifflib.invprobs.VisualScalarOptimizer

com.mockturtlesolutions.snifflib.invprobs.NMSimplex

All Implemented Interfaces:

SimulatedAnnealing

public class NMSimplex
extends VisualScalarOptimizer
implements SimulatedAnnealing

Field Summary

Fields

Modifier and Type	Field and Description
protected int	BestCount
protected java.lang.String	ConvergenceMsg
protected java.lang.String	CurrentMove
protected DblParamSetSimplex	CurrentSimplex
protected int	FunEvals
protected int	Iter
protected int	N
protected NMSimplexConfiguration	Options
protected ParameterValuePair	PrevBest
protected java.lang.String	PrevMove
protected ParameterValuePair	ReflectPt
protected boolean	ScalarMinimization
protected DblMatrix	Temperature
protected ParameterValuePair	TrialPt

Fields inherited from class com.mockturtlesolutions.snifflib.invprobs.VisualScalarOptimizer

InitialGuess, Model, NumParams

Constructor Summary

Constructors

Constructor and Description
NMSimplex(OptimizableScalar Model, DblParamSet InitGuess) Construct a new Nelder-Mead Simplex using an model and an intial parameter guess.

Method Summary

Methods

Modifier and Type	Method and Description
boolean	constraintsOK()
void	ContractInside()
void	ContractOutside()
void	displaySimplex() Display the progress of the simplex.
DblMatrix	evaluateTemperature() Determine the temperature for a simulated annealing schedule.
void	Expand()
DblParamSet	getBestEstimate() Get the estimated parameter value at the estimated parameter value.
DblMatrix	getBestValue() Get the scalar value at the estimated parameter value.
NMSimplexConfiguration	getConfiguration()
DblMatrix	getFitnessValue(DblParamSet param) Get the fitness value for a particular parameter set.
void	getInitialSimplex() Return a DblParamSetSimplex to use as the initial simplex of an optimization.
NMSimplexConfiguration	getOptions()
DblMatrix	getTemperature() Get a temperature for a simulated annealing schedule.
boolean	hasNotConverged()
void	Initial() Initialize all fitness values of the CurrentSimplex.
void	Reflect()
void	setOptions(NMSimplexConfiguration S)
void	setTemperature(DblMatrix temp) Set a temperature for a simulated annealing schedule.
void	Shrink() Shrink the simplex about the average.
void	solve() Minimize until convergence.
void	update() Implementation of the Nelder-Mead Simplex updating rules.
void	visualDiagnostics() Typically this shows visually some convergence diagnostics.
void	visuallyConfigure() Typically this allows setting of optimization settings and preferences related to performance.

Methods inherited from class java.lang.Object

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

Field Detail

CurrentSimplex

protected DblParamSetSimplex CurrentSimplex

Options

protected NMSimplexConfiguration Options

N

protected int N

Iter

protected int Iter

PrevBest

protected ParameterValuePair PrevBest

FunEvals

protected int FunEvals

PrevMove

protected java.lang.String PrevMove

ReflectPt

protected ParameterValuePair ReflectPt

TrialPt

protected ParameterValuePair TrialPt

CurrentMove

protected java.lang.String CurrentMove

Temperature

protected DblMatrix Temperature

ScalarMinimization

protected boolean ScalarMinimization

ConvergenceMsg

protected java.lang.String ConvergenceMsg

BestCount

protected int BestCount

Constructor Detail

NMSimplex

public NMSimplex(OptimizableScalar Model,
         DblParamSet InitGuess)

Construct a new Nelder-Mead Simplex using an model and an intial parameter guess.

Method Detail

getOptions

public NMSimplexConfiguration getOptions()

getBestEstimate

public DblParamSet getBestEstimate()

Get the estimated parameter value at the estimated parameter value.

Specified by:

getBestEstimate in class VisualScalarOptimizer

getBestValue

public DblMatrix getBestValue()

Get the scalar value at the estimated parameter value.

Specified by:

getBestValue in class VisualScalarOptimizer

visuallyConfigure

public void visuallyConfigure()

Typically this allows setting of optimization settings and preferences related to performance.

Specified by:

visuallyConfigure in class VisualScalarOptimizer

visualDiagnostics

public void visualDiagnostics()

Typically this shows visually some convergence diagnostics. Typically no class values are modified by this method.

Specified by:

visualDiagnostics in class VisualScalarOptimizer

Shrink

public void Shrink()

Shrink the simplex about the average.

ContractInside

public void ContractInside()

ContractOutside

public void ContractOutside()

Reflect

public void Reflect()

Expand

public void Expand()

Initial

public void Initial()

Initialize all fitness values of the CurrentSimplex.

hasNotConverged

public boolean hasNotConverged()

constraintsOK

public boolean constraintsOK()

getInitialSimplex

public void getInitialSimplex()

Return a DblParamSetSimplex to use as the initial simplex of an optimization.

getConfiguration

public NMSimplexConfiguration getConfiguration()

setOptions

public void setOptions(NMSimplexConfiguration S)

getFitnessValue

public DblMatrix getFitnessValue(DblParamSet param)

Get the fitness value for a particular parameter set.

update

public void update()

Implementation of the Nelder-Mead Simplex updating rules.

displaySimplex

public void displaySimplex()

Display the progress of the simplex. Displays the current simplex.

solve

public void solve()

Minimize until convergence.

Specified by:

solve in class VisualScalarOptimizer

getTemperature

public DblMatrix getTemperature()

Get a temperature for a simulated annealing schedule.

Specified by:

getTemperature in interface SimulatedAnnealing

setTemperature

public void setTemperature(DblMatrix temp)

Set a temperature for a simulated annealing schedule.

Specified by:

setTemperature in interface SimulatedAnnealing

evaluateTemperature

public DblMatrix evaluateTemperature()

Determine the temperature for a simulated annealing schedule.

Specified by:

evaluateTemperature in interface SimulatedAnnealing