Partial implementation of the Optimizer interface. More...

#include <Bpp/Numeric/Function/AbstractOptimizer.h>

Inheritance diagram for bpp::AbstractOptimizer:

Collaboration diagram for bpp::AbstractOptimizer:

Public Member Functions
	AbstractOptimizer (Function *function=0)

	AbstractOptimizer (const AbstractOptimizer &opt)

AbstractOptimizer &	operator= (const AbstractOptimizer &opt)

virtual	~AbstractOptimizer ()

void	updateParameters (bool yn)
	Tell if we shall update all parameters after one optimization step. More...

bool	updateParameters () const
	Tell if we shall update all parameters after one optimization step. More...

void	setOptimizationProgressCharacter (const std::string &c)
	Set the character to be displayed during optimization. More...

const std::string &	getOptimizationProgressCharacter () const

Optimizer *	clone () const =0
	Create a copy of this object and send a pointer to it. More...

The Optimizer interface.
void	init (const ParameterList &params)
	Basic implementation. More...

double	step ()
	Basic implementation. More...

double	optimize ()
	Basic implementation. More...

bool	isInitialized () const

const ParameterList &	getParameters () const

double	getParameterValue (const std::string &name) const

void	setFunction (Function *function)
	Set the function to optimize. More...

const Function *	getFunction () const
	Get the current function being optimized. More...

Function *	getFunction ()
	Get the current function being optimized. More...

bool	hasFunction () const
	Tell if a funciton is associatied to this optimizer. More...

double	getFunctionValue () const
	Get the current function value. More...

void	setMessageHandler (OutputStream *mh)
	Set the message handler for this optimizer. More...

OutputStream *	getMessageHandler () const

void	setProfiler (OutputStream *profiler)
	Set the profiler for this optimizer. More...

OutputStream *	getProfiler () const

unsigned int	getNumberOfEvaluations () const
	Get the number of function evaluations performed since the call of the init function. More...

void	setStopCondition (const OptimizationStopCondition &stopCondition)
	Set the stop condition of the optimization algorithm. More...

OptimizationStopCondition *	getStopCondition ()
	Get the stop condition of the optimization algorithm. More...

const OptimizationStopCondition *	getStopCondition () const
	Get the stop condition of the optimization algorithm. More...

OptimizationStopCondition *	getDefaultStopCondition ()
	Get the default stop condition of the optimization algorithm. More...

const OptimizationStopCondition *	getDefaultStopCondition () const
	Get the default stop condition of the optimization algorithm. More...

bool	isToleranceReached () const
	Tell if the tolerance level is reached. More...

bool	isMaximumNumberOfEvaluationsReached () const
	Tell if the maximum number of function evaluations is reached. More...

void	setMaximumNumberOfEvaluations (unsigned int max)
	Set the maximum number of function evaluation to perform during optimization. More...

void	setVerbose (unsigned int v)
	Set the verbose level. More...

unsigned int	getVerbose () const
	Get the verbose level. More...

void	setConstraintPolicy (const std::string &constraintPolicy)
	Set the constraint policy for this optimizer. More...

std::string	getConstraintPolicy () const
	Get the constraint policy for this optimizer. More...

void	addOptimizationListener (OptimizationListener *listener)
	Register a listener to this class. More...

Protected Member Functions
virtual void	doInit (const ParameterList &params)=0
	This function is called by the init() method and contains all calculations. More...

virtual double	doStep ()=0
	This function is called by the step() method and contains all calculations. More...

ParameterList &	getParameters_ ()

Parameter &	getParameter_ (size_t i)

Function *	getFunction_ ()

void	setDefaultStopCondition_ (OptimizationStopCondition *osc)

Inner utilitary functions
void	autoParameter ()
	Build a list of AutoParameter instead of Parameter. More...

void	ignoreConstraints ()
	Remove the constraints of all the arguments. More...

void	profile (double v)
	Print to the profile if there is one. More...

void	profile (unsigned int v)
	Print to the profile if there is one. More...

void	profile (const std::string &s)
	Print to the profile if there is one. More...

void	profileln (double v)
	Print to the profile if there is one and end line. More...

void	profileln (unsigned int v)
	Print to the profile if there is one and end line. More...

void	profileln (const std::string &s)
	Print to the profile if there is one and end line. More...

void	printPoint (const ParameterList &params, double value)
	Print parameters and corresponding function evaluation to profiler. More...

void	printMessage (const std::string &message)
	Give a message to print to the message handler. More...

void	fireOptimizationInitializationPerformed (const OptimizationEvent &event)
	Notify all listeners that optimizer initialization was performed. More...

void	fireOptimizationStepPerformed (const OptimizationEvent &event)
	Notify all listeners that an optimization step was performed. More...

bool	listenerModifiesParameters () const

Protected Attributes
unsigned int	nbEvalMax_
	The maximum number of function evaluations allowed. More...

unsigned int	nbEval_
	The current number of function evaluations achieved. More...

double	currentValue_
	The current value of the function. More...

bool	tolIsReached_
	Tell if the tolerance level has been reached. More...

Private Attributes
Function *	function_
	The function to optimize. More...

ParameterList	parameters_
	The parameters that will be optimized. More...

OutputStream *	messageHandler_
	The message handler. More...

OutputStream *	profiler_
	The profiler. More...

std::string	constraintPolicy_
	The constraint policy. More...

OptimizationStopCondition *	stopCondition_
	The stoping condition to use while optimizing. More...

OptimizationStopCondition *	defaultStopCondition_
	The default stoping condition to use while optimizing. More...

unsigned int	verbose_
	State of the verbose mode: > 0 = enabled. More...

bool	isInitialized_
	Check if the optimizer have been feeded with initial parameters values. More...

time_t	startTime_

std::vector< OptimizationListener * >	listeners_

bool	updateParameters_

std::string	stepChar_

Detailed Description

Partial implementation of the Optimizer interface.

This implementation is designed for unconstrained or simple-bounded optimization. You should not use it with global contraints. It also enables the gestion of listeners by maintaining a vector of pointers toward the listener. Important note: this list of listener is not duplicated in cas of copy of the Optimizer, as listeners are expected to be bounded ot a particular instance.:if expand("%") == ""|browse confirm w|else|confirm w|endif

Definition at line 60 of file AbstractOptimizer.h.

Constructor & Destructor Documentation

◆ AbstractOptimizer() [1/2]

AbstractOptimizer::AbstractOptimizer ( Function * function = 0 )

Definition at line 56 of file AbstractOptimizer.cpp.

◆ AbstractOptimizer() [2/2]

AbstractOptimizer::AbstractOptimizer ( const AbstractOptimizer & opt )

Definition at line 71 of file AbstractOptimizer.cpp.

References autoParameter(), bpp::OptimizationStopCondition::clone(), constraintPolicy_, bpp::AutoParameter::CONSTRAINTS_AUTO, bpp::AutoParameter::CONSTRAINTS_IGNORE, defaultStopCondition_, ignoreConstraints(), isInitialized_, bpp::OptimizationStopCondition::setOptimizer(), and stopCondition_.

◆ ~AbstractOptimizer()

virtual bpp::AbstractOptimizer::~AbstractOptimizer ( )

inlinevirtual

Definition at line 157 of file AbstractOptimizer.h.

References defaultStopCondition_, and stopCondition_.

Member Function Documentation

◆ addOptimizationListener()

void bpp::AbstractOptimizer::addOptimizationListener ( OptimizationListener * listener )

inlinevirtual

Register a listener to this class.

All registered listeners will be informed when an optimization event occur. See the documentation of the class to know what kind of events are supported.

Parameters

listener A listener to be registered with.

Implements bpp::Optimizer.

Definition at line 225 of file AbstractOptimizer.h.

References listeners_.

◆ autoParameter()

void AbstractOptimizer::autoParameter ( )

protected

Build a list of AutoParameter instead of Parameter.

Definition at line 306 of file AbstractOptimizer.cpp.

References messageHandler_, parameters_, bpp::AutoParameter::setMessageHandler(), bpp::ParameterList::setParameter(), and bpp::ParameterList::size().

Referenced by AbstractOptimizer(), init(), and operator=().

◆ clone()

Optimizer* bpp::Optimizer::clone ( ) const

pure virtualinherited

Create a copy of this object and send a pointer to it.

Returns: A pointer toward the copy object.

Implements bpp::Clonable.

Implemented in bpp::SimpleNewtonMultiDimensions, bpp::SimpleMultiDimensions, bpp::PowellMultiDimensions, bpp::NewtonOneDimension, bpp::NewtonBacktrackOneDimension, bpp::MetaOptimizer, bpp::GoldenSectionSearch, bpp::DownhillSimplexMethod, bpp::ConjugateGradientMultiDimensions, bpp::BrentOneDimension, and bpp::BfgsMultiDimensions.

◆ doInit()

virtual void bpp::AbstractOptimizer::doInit ( const ParameterList & params )

protectedpure virtual

This function is called by the init() method and contains all calculations.

Parameters

params The parameters to use for initialization.

Implemented in bpp::SimpleNewtonMultiDimensions, bpp::SimpleMultiDimensions, bpp::PowellMultiDimensions, bpp::NewtonOneDimension, bpp::NewtonBacktrackOneDimension, bpp::GoldenSectionSearch, bpp::DownhillSimplexMethod, bpp::ConjugateGradientMultiDimensions, bpp::BrentOneDimension, bpp::BfgsMultiDimensions, and bpp::MetaOptimizer.

Referenced by init().

◆ doStep()

virtual double bpp::AbstractOptimizer::doStep ( )

protectedpure virtual

This function is called by the step() method and contains all calculations.

Returns: The value of the function after the optimization step.

Implemented in bpp::SimpleNewtonMultiDimensions, bpp::SimpleMultiDimensions, bpp::PowellMultiDimensions, bpp::NewtonOneDimension, bpp::NewtonBacktrackOneDimension, bpp::MetaOptimizer, bpp::GoldenSectionSearch, bpp::DownhillSimplexMethod, bpp::ConjugateGradientMultiDimensions, bpp::BrentOneDimension, and bpp::BfgsMultiDimensions.

Referenced by step().

◆ fireOptimizationInitializationPerformed()

void AbstractOptimizer::fireOptimizationInitializationPerformed ( const OptimizationEvent & event )

protected

Notify all listeners that optimizer initialization was performed.

This method should be called by the init method.

Parameters

event An OptimizationEvent object.

Definition at line 328 of file AbstractOptimizer.cpp.

References listeners_.

Referenced by init().

◆ fireOptimizationStepPerformed()

void AbstractOptimizer::fireOptimizationStepPerformed ( const OptimizationEvent & event )

protected

Notify all listeners that an optimization step was performed.

This method should be called by the step method.

Parameters

event An OptimizationEvent object.

Definition at line 338 of file AbstractOptimizer.cpp.

References listeners_.

Referenced by step().

◆ getConstraintPolicy()

std::string bpp::AbstractOptimizer::getConstraintPolicy ( ) const

inlinevirtual

Get the constraint policy for this optimizer.

Returns: The constraint policy.

Implements bpp::Optimizer.

Definition at line 224 of file AbstractOptimizer.h.

References constraintPolicy_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::SimpleMultiDimensions::doInit(), and bpp::SimpleNewtonMultiDimensions::doInit().

◆ getDefaultStopCondition() [1/2]

OptimizationStopCondition* bpp::AbstractOptimizer::getDefaultStopCondition ( )

inlinevirtual

Get the default stop condition of the optimization algorithm.

Returns: The default stop condition used while optimizing.

Implements bpp::Optimizer.

Definition at line 216 of file AbstractOptimizer.h.

References defaultStopCondition_.

Referenced by bpp::BfgsMultiDimensions::BfgsMultiDimensions(), bpp::BrentOneDimension::BrentOneDimension(), bpp::ConjugateGradientMultiDimensions::ConjugateGradientMultiDimensions(), bpp::DownhillSimplexMethod::DownhillSimplexMethod(), bpp::GoldenSectionSearch::GoldenSectionSearch(), bpp::MetaOptimizer::MetaOptimizer(), bpp::NewtonBacktrackOneDimension::NewtonBacktrackOneDimension(), bpp::NewtonOneDimension::NewtonOneDimension(), bpp::PowellMultiDimensions::PowellMultiDimensions(), bpp::SimpleMultiDimensions::SimpleMultiDimensions(), and bpp::SimpleNewtonMultiDimensions::SimpleNewtonMultiDimensions().

◆ getDefaultStopCondition() [2/2]

const OptimizationStopCondition* bpp::AbstractOptimizer::getDefaultStopCondition ( ) const

inlinevirtual

Get the default stop condition of the optimization algorithm.

Returns: The default stop condition used while optimizing.

Implements bpp::Optimizer.

Definition at line 217 of file AbstractOptimizer.h.

References defaultStopCondition_.

◆ getFunction() [1/2]

Function* bpp::AbstractOptimizer::getFunction ( )

inlinevirtual

Get the current function being optimized.

Returns: A const pointer toward the function being optimized.

Implements bpp::Optimizer.

Reimplemented in bpp::NewtonOneDimension, bpp::ConjugateGradientMultiDimensions, and bpp::BfgsMultiDimensions.

Definition at line 196 of file AbstractOptimizer.h.

References function_.

◆ getFunction() [2/2]

const Function* bpp::AbstractOptimizer::getFunction ( ) const

inlinevirtual

Get the current function being optimized.

Returns: A const pointer toward the function being optimized.

Implements bpp::Optimizer.

Reimplemented in bpp::NewtonOneDimension, bpp::ConjugateGradientMultiDimensions, and bpp::BfgsMultiDimensions.

Definition at line 195 of file AbstractOptimizer.h.

References function_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::BrentOneDimension::doInit(), bpp::DownhillSimplexMethod::doInit(), bpp::GoldenSectionSearch::doInit(), bpp::NewtonBacktrackOneDimension::doInit(), bpp::PowellMultiDimensions::doInit(), bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::BrentOneDimension::doStep(), bpp::DownhillSimplexMethod::doStep(), bpp::GoldenSectionSearch::doStep(), bpp::MetaOptimizer::doStep(), bpp::NewtonBacktrackOneDimension::doStep(), bpp::PowellMultiDimensions::doStep(), bpp::SimpleMultiDimensions::doStep(), bpp::SimpleNewtonMultiDimensions::doStep(), bpp::BfgsMultiDimensions::getFunction(), bpp::ConjugateGradientMultiDimensions::getFunction(), bpp::NewtonOneDimension::getFunction(), bpp::BrentOneDimension::optimize(), bpp::DownhillSimplexMethod::optimize(), bpp::PowellMultiDimensions::optimize(), and bpp::DownhillSimplexMethod::tryExtrapolation().

◆ getFunction_()

Function* bpp::AbstractOptimizer::getFunction_ ( )

inlineprotected

Definition at line 377 of file AbstractOptimizer.h.

References function_.

Referenced by bpp::BfgsMultiDimensions::getFunction_(), bpp::ConjugateGradientMultiDimensions::getFunction_(), bpp::NewtonBacktrackOneDimension::getFunction_(), and bpp::NewtonOneDimension::getFunction_().

◆ getFunctionValue()

double bpp::AbstractOptimizer::getFunctionValue ( ) const

inlinevirtual

Get the current function value.

Returns: The value of the function at the point specified by _parameters.

Exceptions

NullPointerException If no function is associated with this optimizer.

Implements bpp::Optimizer.

Reimplemented in bpp::GoldenSectionSearch.

Definition at line 198 of file AbstractOptimizer.h.

References currentValue_, and function_.

◆ getMessageHandler()

OutputStream* bpp::AbstractOptimizer::getMessageHandler ( ) const

inlinevirtual

Returns: The stream used for handling messages, if any.

Implements bpp::Optimizer.

Definition at line 205 of file AbstractOptimizer.h.

References messageHandler_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), and bpp::PowellMultiDimensions::doStep().

◆ getNumberOfEvaluations()

unsigned int bpp::AbstractOptimizer::getNumberOfEvaluations ( ) const

inlinevirtual

Get the number of function evaluations performed since the call of the init function.

Returns: The number of function evaluations.

Implements bpp::Optimizer.

Definition at line 209 of file AbstractOptimizer.h.

References nbEval_.

Referenced by bpp::SimpleMultiDimensions::doStep(), bpp::SimpleNewtonMultiDimensions::doStep(), bpp::OneDimensionOptimizationTools::lineMinimization(), and bpp::OneDimensionOptimizationTools::lineSearch().

◆ getOptimizationProgressCharacter()

const std::string& bpp::AbstractOptimizer::getOptimizationProgressCharacter ( ) const

inline

Returns: The character to be displayed during optimization.

Definition at line 263 of file AbstractOptimizer.h.

References stepChar_.

◆ getParameter_()

Parameter& bpp::AbstractOptimizer::getParameter_ ( size_t i )

inlineprotected

Definition at line 376 of file AbstractOptimizer.h.

References parameters_.

Referenced by bpp::BrentOneDimension::doInit(), bpp::GoldenSectionSearch::doInit(), bpp::BrentOneDimension::doStep(), bpp::GoldenSectionSearch::doStep(), and bpp::NewtonBacktrackOneDimension::doStep().

◆ getParameters()

const ParameterList& bpp::AbstractOptimizer::getParameters ( ) const

inlinevirtual

Returns: The parameters with their current values.

Implements bpp::Optimizer.

Definition at line 188 of file AbstractOptimizer.h.

References parameters_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::BfgsMultiDimensions::doInit(), bpp::BrentOneDimension::doInit(), bpp::DownhillSimplexMethod::doInit(), bpp::GoldenSectionSearch::doInit(), bpp::NewtonBacktrackOneDimension::doInit(), bpp::NewtonOneDimension::doInit(), bpp::PowellMultiDimensions::doInit(), bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::BfgsMultiDimensions::doStep(), bpp::BrentOneDimension::doStep(), bpp::ConjugateGradientMultiDimensions::doStep(), bpp::GoldenSectionSearch::doStep(), bpp::MetaOptimizer::doStep(), bpp::NewtonBacktrackOneDimension::doStep(), bpp::NewtonOneDimension::doStep(), bpp::PowellMultiDimensions::doStep(), bpp::SimpleMultiDimensions::doStep(), bpp::SimpleNewtonMultiDimensions::doStep(), bpp::BfgsMultiDimensions::getGradient(), bpp::ConjugateGradientMultiDimensions::getGradient(), bpp::DownhillSimplexMethod::getPSum(), bpp::BrentOneDimension::optimize(), bpp::PowellMultiDimensions::optimize(), bpp::BfgsMultiDimensions::setDirection(), and bpp::DownhillSimplexMethod::tryExtrapolation().

◆ getParameters_()

ParameterList& bpp::AbstractOptimizer::getParameters_ ( )

inlineprotected

Definition at line 375 of file AbstractOptimizer.h.

References parameters_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::BfgsMultiDimensions::doStep(), bpp::ConjugateGradientMultiDimensions::doStep(), bpp::DownhillSimplexMethod::doStep(), bpp::MetaOptimizer::doStep(), bpp::NewtonOneDimension::doStep(), bpp::PowellMultiDimensions::doStep(), bpp::SimpleMultiDimensions::doStep(), and bpp::SimpleNewtonMultiDimensions::doStep().

◆ getParameterValue()

double bpp::AbstractOptimizer::getParameterValue ( const std::string & name ) const

inlinevirtual

Returns: The value associated with a given parameter name.

Implements bpp::Optimizer.

Definition at line 189 of file AbstractOptimizer.h.

References bpp::ParameterList::getParameterValue(), and parameters_.

◆ getProfiler()

OutputStream* bpp::AbstractOptimizer::getProfiler ( ) const

inlinevirtual

Returns: The stream used for profiling, if any.

Implements bpp::Optimizer.

Definition at line 207 of file AbstractOptimizer.h.

References profiler_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::SimpleMultiDimensions::doInit(), and bpp::SimpleNewtonMultiDimensions::doInit().

◆ getStopCondition() [1/2]

OptimizationStopCondition* bpp::AbstractOptimizer::getStopCondition ( )

inlinevirtual

Get the stop condition of the optimization algorithm.

Returns: The stop condition used while optimizing.

Implements bpp::Optimizer.

Definition at line 214 of file AbstractOptimizer.h.

References stopCondition_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::NewtonBacktrackOneDimension::doInit(), bpp::NewtonOneDimension::doInit(), bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::BrentOneDimension::doStep(), bpp::ConjugateGradientMultiDimensions::doStep(), bpp::GoldenSectionSearch::doStep(), bpp::MetaOptimizer::doStep(), bpp::PowellMultiDimensions::doStep(), bpp::SimpleMultiDimensions::doStep(), bpp::OneDimensionOptimizationTools::lineMinimization(), bpp::OneDimensionOptimizationTools::lineSearch(), bpp::MetaOptimizer::MetaOptimizer(), and bpp::SimpleMultiDimensions::setFunction().

◆ getStopCondition() [2/2]

const OptimizationStopCondition* bpp::AbstractOptimizer::getStopCondition ( ) const

inlinevirtual

Get the stop condition of the optimization algorithm.

Returns: The stop condition used while optimizing.

Implements bpp::Optimizer.

Definition at line 215 of file AbstractOptimizer.h.

References stopCondition_.

◆ getVerbose()

unsigned int bpp::AbstractOptimizer::getVerbose ( ) const

inlinevirtual

Get the verbose level.

Returns: verbose level.

Implements bpp::Optimizer.

Definition at line 222 of file AbstractOptimizer.h.

References verbose_.

Referenced by bpp::MetaOptimizer::doInit(), bpp::BrentOneDimension::doInit(), bpp::GoldenSectionSearch::doInit(), bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::BfgsMultiDimensions::doStep(), bpp::ConjugateGradientMultiDimensions::doStep(), bpp::MetaOptimizer::doStep(), bpp::PowellMultiDimensions::doStep(), bpp::SimpleMultiDimensions::doStep(), and bpp::SimpleNewtonMultiDimensions::doStep().

◆ hasFunction()

bool bpp::AbstractOptimizer::hasFunction ( ) const

inlinevirtual

Tell if a funciton is associatied to this optimizer.

Returns: True if a function has been associated to this optimizer.

Implements bpp::Optimizer.

Definition at line 197 of file AbstractOptimizer.h.

References function_.

Referenced by bpp::GoldenSectionSearch::getFunctionValue().

◆ ignoreConstraints()

void AbstractOptimizer::ignoreConstraints ( )

protected

Remove the constraints of all the arguments.

Definition at line 318 of file AbstractOptimizer.cpp.

References parameters_, and bpp::ParameterList::size().

Referenced by AbstractOptimizer(), init(), and operator=().

◆ init()

void AbstractOptimizer::init ( const ParameterList & params )

virtual

Basic implementation.

Store all parameters, call the doInit method, print to profiler, initialize timer and notify all listeners.

Implements bpp::Optimizer.

Definition at line 158 of file AbstractOptimizer.cpp.

References autoParameter(), constraintPolicy_, bpp::AutoParameter::CONSTRAINTS_AUTO, bpp::AutoParameter::CONSTRAINTS_IGNORE, currentValue_, doInit(), fireOptimizationInitializationPerformed(), function_, bpp::Function::getValue(), ignoreConstraints(), bpp::OptimizationStopCondition::init(), isInitialized_, nbEval_, parameters_, printPoint(), profile(), profileln(), bpp::ParameterList::size(), startTime_, stopCondition_, and tolIsReached_.

Referenced by bpp::SimpleMultiDimensions::doStep(), bpp::SimpleNewtonMultiDimensions::doStep(), bpp::OneDimensionOptimizationTools::lineMinimization(), and bpp::OneDimensionOptimizationTools::lineSearch().

◆ isInitialized()

bool bpp::AbstractOptimizer::isInitialized ( ) const

inlinevirtual

Returns: 'true' if this optimizer has been initialized.

Implements bpp::Optimizer.

Definition at line 187 of file AbstractOptimizer.h.

References isInitialized_.

◆ isMaximumNumberOfEvaluationsReached()

bool bpp::AbstractOptimizer::isMaximumNumberOfEvaluationsReached ( ) const

inlinevirtual

Tell if the maximum number of function evaluations is reached.

Returns: Whether the maximum number of function evaluations is reached or not.

Implements bpp::Optimizer.

Definition at line 219 of file AbstractOptimizer.h.

References nbEval_, and nbEvalMax_.

◆ isToleranceReached()

bool bpp::AbstractOptimizer::isToleranceReached ( ) const

inlinevirtual

Tell if the tolerance level is reached.

Returns: Whether the tolerance is reached or not.

See also: OptimizationStopCondition

Implements bpp::Optimizer.

Definition at line 218 of file AbstractOptimizer.h.

References tolIsReached_.

◆ listenerModifiesParameters()

bool AbstractOptimizer::listenerModifiesParameters ( ) const

protected

Definition at line 348 of file AbstractOptimizer.cpp.

References listeners_.

Referenced by step().

◆ operator=()

AbstractOptimizer & AbstractOptimizer::operator= ( const AbstractOptimizer & opt )

◆ optimize()

double AbstractOptimizer::optimize ( )

virtual

Basic implementation.

Call the step method untill tolerance is reached.

Implements bpp::Optimizer.

Reimplemented in bpp::PowellMultiDimensions, bpp::DownhillSimplexMethod, and bpp::BrentOneDimension.

Definition at line 216 of file AbstractOptimizer.cpp.

References currentValue_, bpp::ApplicationTools::displayUnlimitedGauge(), isInitialized_, nbEval_, nbEvalMax_, step(), tolIsReached_, and verbose_.

Referenced by bpp::SimpleNewtonMultiDimensions::doStep(), bpp::OneDimensionOptimizationTools::lineSearch(), bpp::BrentOneDimension::optimize(), bpp::DownhillSimplexMethod::optimize(), and bpp::PowellMultiDimensions::optimize().

◆ printMessage()

void AbstractOptimizer::printMessage ( const std::string & message )

protected

Give a message to print to the message handler.

Parameters

message The message to print.

Definition at line 298 of file AbstractOptimizer.cpp.

References messageHandler_.

Referenced by bpp::BrentOneDimension::doInit(), bpp::GoldenSectionSearch::doInit(), bpp::BfgsMultiDimensions::doStep(), and bpp::NewtonOneDimension::doStep().

◆ printPoint()

void AbstractOptimizer::printPoint	(	const ParameterList &	params,
		double	value
	)

protected

Print parameters and corresponding function evaluation to profiler.

Parameters

params	The parameters to print.
value	The function evaluation.

Definition at line 279 of file AbstractOptimizer.cpp.

References nbEval_, profile(), profileln(), bpp::ParameterList::size(), startTime_, and bpp::TextTools::toString().

Referenced by bpp::PowellMultiDimensions::doStep(), init(), and step().

◆ profile() [1/3]

void AbstractOptimizer::profile ( const std::string & s )

protected

Print to the profile if there is one.

Parameters

s	The string to print to the profile.

Definition at line 263 of file AbstractOptimizer.cpp.

References profiler_.

◆ profile() [2/3]

void AbstractOptimizer::profile ( double v )

protected

Print to the profile if there is one.

Parameters

v	The double value to print.

Definition at line 232 of file AbstractOptimizer.cpp.

References profiler_.

Referenced by init(), and printPoint().

◆ profile() [3/3]

void AbstractOptimizer::profile ( unsigned int v )

protected

Print to the profile if there is one.

Parameters

v	The unsigned int value to print.

Definition at line 248 of file AbstractOptimizer.cpp.

References profiler_.

◆ profileln() [1/3]

void AbstractOptimizer::profileln ( const std::string & s )

protected

Print to the profile if there is one and end line.

Parameters

s	The string to print to the profile.

Definition at line 271 of file AbstractOptimizer.cpp.

References profiler_.

◆ profileln() [2/3]

void AbstractOptimizer::profileln ( double v )

protected

Print to the profile if there is one and end line.

Parameters

v	The double value to print.

Definition at line 240 of file AbstractOptimizer.cpp.

References profiler_.

Referenced by init(), and printPoint().

◆ profileln() [3/3]

void AbstractOptimizer::profileln ( unsigned int v )

protected

Print to the profile if there is one and end line.

Parameters

v	The unsigned int value to print.

Definition at line 255 of file AbstractOptimizer.cpp.

References profiler_.

◆ setConstraintPolicy()

void bpp::AbstractOptimizer::setConstraintPolicy ( const std::string & constraintPolicy )

inlinevirtual

Set the constraint policy for this optimizer.

Parameters

constraintPolicy The constraint policy.

Implements bpp::Optimizer.

Definition at line 223 of file AbstractOptimizer.h.

References constraintPolicy_.

Referenced by bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::OneDimensionOptimizationTools::lineMinimization(), and bpp::OneDimensionOptimizationTools::lineSearch().

◆ setDefaultStopCondition_()

void bpp::AbstractOptimizer::setDefaultStopCondition_ ( OptimizationStopCondition * osc )

inlineprotected

Definition at line 378 of file AbstractOptimizer.h.

References defaultStopCondition_.

Referenced by bpp::BfgsMultiDimensions::BfgsMultiDimensions(), bpp::BrentOneDimension::BrentOneDimension(), bpp::ConjugateGradientMultiDimensions::ConjugateGradientMultiDimensions(), bpp::DownhillSimplexMethod::DownhillSimplexMethod(), bpp::GoldenSectionSearch::GoldenSectionSearch(), bpp::MetaOptimizer::MetaOptimizer(), bpp::NewtonBacktrackOneDimension::NewtonBacktrackOneDimension(), bpp::NewtonOneDimension::NewtonOneDimension(), bpp::PowellMultiDimensions::PowellMultiDimensions(), bpp::SimpleMultiDimensions::SimpleMultiDimensions(), and bpp::SimpleNewtonMultiDimensions::SimpleNewtonMultiDimensions().

◆ setFunction()

void bpp::AbstractOptimizer::setFunction ( Function * function )

inlinevirtual

Set the function to optimize.

Parameters

function The function to optimize.

Implements bpp::Optimizer.

Reimplemented in bpp::SimpleNewtonMultiDimensions, bpp::SimpleMultiDimensions, and bpp::MetaOptimizer.

Definition at line 190 of file AbstractOptimizer.h.

References function_, bpp::OptimizationStopCondition::init(), and stopCondition_.

Referenced by bpp::MetaOptimizer::setFunction(), bpp::SimpleMultiDimensions::setFunction(), and bpp::SimpleNewtonMultiDimensions::setFunction().

◆ setMaximumNumberOfEvaluations()

void bpp::AbstractOptimizer::setMaximumNumberOfEvaluations ( unsigned int max )

inlinevirtual

Set the maximum number of function evaluation to perform during optimization.

Parameters

max	The maximum number of evaluations to perform.

Implements bpp::Optimizer.

Definition at line 220 of file AbstractOptimizer.h.

References nbEvalMax_.

Referenced by bpp::BrentOneDimension::BrentOneDimension(), bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), and bpp::NewtonBacktrackOneDimension::NewtonBacktrackOneDimension().

◆ setMessageHandler()

void bpp::AbstractOptimizer::setMessageHandler ( OutputStream * mh )

inlinevirtual

Set the message handler for this optimizer.

The message handler keeps all messages that the optimizer may send. The default handler is set to standard output, but you can pass any ostream object (cerr, ofstream, etc.).

A NULL pointer disables message output.

Parameters

mh	The message handler to use.

Implements bpp::Optimizer.

Definition at line 204 of file AbstractOptimizer.h.

References messageHandler_.

Referenced by bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::OneDimensionOptimizationTools::lineMinimization(), and bpp::OneDimensionOptimizationTools::lineSearch().

◆ setOptimizationProgressCharacter()

void bpp::AbstractOptimizer::setOptimizationProgressCharacter ( const std::string & c )

inline

Set the character to be displayed during optimization.

Parameters

c	A character.

Definition at line 259 of file AbstractOptimizer.h.

References stepChar_.

Referenced by bpp::BfgsMultiDimensions::BfgsMultiDimensions(), bpp::OneDimensionOptimizationTools::lineMinimization(), bpp::OneDimensionOptimizationTools::lineSearch(), bpp::MetaOptimizer::MetaOptimizer(), bpp::SimpleMultiDimensions::SimpleMultiDimensions(), and bpp::SimpleNewtonMultiDimensions::SimpleNewtonMultiDimensions().

◆ setProfiler()

void bpp::AbstractOptimizer::setProfiler ( OutputStream * profiler )

inlinevirtual

Set the profiler for this optimizer.

The profiler keeps all the intermediate values taken by the parameters. The default profiler is set to standard output, but you can pass any ostream object (cerr, ofstream, etc.).

A NULL pointer disables message output.

Parameters

profiler The profiler to use.

Implements bpp::Optimizer.

Definition at line 206 of file AbstractOptimizer.h.

References profiler_.

Referenced by bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::OneDimensionOptimizationTools::lineMinimization(), and bpp::OneDimensionOptimizationTools::lineSearch().

◆ setStopCondition()

void bpp::AbstractOptimizer::setStopCondition ( const OptimizationStopCondition & stopCondition )

inlinevirtual

Set the stop condition of the optimization algorithm.

Parameters

stopCondition The stop condition to use while optimizing.

See also: OptimizationStopCondition.

Implements bpp::Optimizer.

Definition at line 210 of file AbstractOptimizer.h.

References bpp::OptimizationStopCondition::clone(), and stopCondition_.

Referenced by bpp::BfgsMultiDimensions::BfgsMultiDimensions(), bpp::BrentOneDimension::BrentOneDimension(), bpp::ConjugateGradientMultiDimensions::ConjugateGradientMultiDimensions(), bpp::DownhillSimplexMethod::DownhillSimplexMethod(), bpp::GoldenSectionSearch::GoldenSectionSearch(), bpp::MetaOptimizer::MetaOptimizer(), bpp::NewtonBacktrackOneDimension::NewtonBacktrackOneDimension(), bpp::NewtonOneDimension::NewtonOneDimension(), bpp::PowellMultiDimensions::PowellMultiDimensions(), bpp::SimpleMultiDimensions::SimpleMultiDimensions(), and bpp::SimpleNewtonMultiDimensions::SimpleNewtonMultiDimensions().

◆ setVerbose()

void bpp::AbstractOptimizer::setVerbose ( unsigned int v )

inlinevirtual

Set the verbose level.

0 = off 1 = on 2 = more verbose 3 = even more, etc.

In most cases, only the 0 and 1 levels are implemented.

Parameters

v	verbose level.

Implements bpp::Optimizer.

Definition at line 221 of file AbstractOptimizer.h.

References verbose_.

Referenced by bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::OneDimensionOptimizationTools::lineMinimization(), and bpp::OneDimensionOptimizationTools::lineSearch().

◆ step()

double AbstractOptimizer::step ( )

virtual

Basic implementation.

Check if the optimizer is initialized, check if parameters need update because of listeners, call the doStep method, print the current point to the profiler, notify all listeners and return the current value of the function.

Implements bpp::Optimizer.

Definition at line 196 of file AbstractOptimizer.cpp.

References currentValue_, doStep(), bpp::Function::f(), fireOptimizationStepPerformed(), function_, bpp::Parametrizable::getParameters(), bpp::OptimizationStopCondition::isToleranceReached(), listenerModifiesParameters(), bpp::ParameterList::matchParametersValues(), parameters_, printPoint(), stopCondition_, tolIsReached_, and updateParameters_.

Referenced by optimize().

◆ updateParameters() [1/2]

bool bpp::AbstractOptimizer::updateParameters ( ) const

inline

Tell if we shall update all parameters after one optimization step.

This is required only for functions that have non-independent parameters, which means that setting one parameter value may modify one or several other parameters. Depending on the optimizer, this may have no effect.

Returns: yn true/false

Definition at line 252 of file AbstractOptimizer.h.

References updateParameters_.

Referenced by bpp::MetaOptimizer::doInit().

◆ updateParameters() [2/2]

void bpp::AbstractOptimizer::updateParameters ( bool yn )

inline

Tell if we shall update all parameters after one optimization step.

This is required only for functions that have non-independent parameters, which means that setting one parameter value may modify one or several other parameters. Depending on the optimizer, this may have no effect.

Parameters

yn	true/false

Definition at line 241 of file AbstractOptimizer.h.

References updateParameters_.

Member Data Documentation

◆ constraintPolicy_

std::string bpp::AbstractOptimizer::constraintPolicy_

private

The constraint policy.

Must be one the following:

CONSTRAINTS_KEEP: keep the constraint associated to the parameters (default).
CONSTRAINTS_IGNORE: remove all constraints.
CONSTRAINTS_AUTO: use AutoParameters to deal with constraints.

See also: AutoParameter

Definition at line 94 of file AbstractOptimizer.h.

Referenced by AbstractOptimizer(), getConstraintPolicy(), init(), operator=(), and setConstraintPolicy().

◆ currentValue_

double bpp::AbstractOptimizer::currentValue_

protected

The current value of the function.

Definition at line 140 of file AbstractOptimizer.h.

Referenced by bpp::NewtonOneDimension::doInit(), bpp::BfgsMultiDimensions::doStep(), bpp::NewtonOneDimension::doStep(), getFunctionValue(), bpp::GoldenSectionSearch::getFunctionValue(), init(), optimize(), bpp::BrentOneDimension::optimize(), and step().

◆ defaultStopCondition_

OptimizationStopCondition* bpp::AbstractOptimizer::defaultStopCondition_

private

The default stoping condition to use while optimizing.

Definition at line 104 of file AbstractOptimizer.h.

Referenced by AbstractOptimizer(), getDefaultStopCondition(), operator=(), setDefaultStopCondition_(), and ~AbstractOptimizer().

◆ function_

Function* bpp::AbstractOptimizer::function_

private

The function to optimize.

Definition at line 67 of file AbstractOptimizer.h.

Referenced by getFunction(), getFunction_(), getFunctionValue(), hasFunction(), init(), operator=(), setFunction(), and step().

◆ isInitialized_

bool bpp::AbstractOptimizer::isInitialized_

private

Check if the optimizer have been feeded with initial parameters values.

Definition at line 116 of file AbstractOptimizer.h.

Referenced by AbstractOptimizer(), init(), isInitialized(), operator=(), and optimize().

◆ listeners_

std::vector<OptimizationListener*> bpp::AbstractOptimizer::listeners_

private

Definition at line 120 of file AbstractOptimizer.h.

Referenced by addOptimizationListener(), fireOptimizationInitializationPerformed(), fireOptimizationStepPerformed(), listenerModifiesParameters(), and operator=().

◆ messageHandler_

OutputStream* bpp::AbstractOptimizer::messageHandler_

private

The message handler.

Definition at line 77 of file AbstractOptimizer.h.

Referenced by autoParameter(), getMessageHandler(), operator=(), printMessage(), and setMessageHandler().

◆ nbEval_

unsigned int bpp::AbstractOptimizer::nbEval_

protected

The current number of function evaluations achieved.

Definition at line 135 of file AbstractOptimizer.h.

Referenced by bpp::DownhillSimplexMethod::doInit(), bpp::BfgsMultiDimensions::doStep(), bpp::ConjugateGradientMultiDimensions::doStep(), bpp::DownhillSimplexMethod::doStep(), bpp::GoldenSectionSearch::doStep(), bpp::MetaOptimizer::doStep(), bpp::PowellMultiDimensions::doStep(), bpp::SimpleMultiDimensions::doStep(), bpp::SimpleNewtonMultiDimensions::doStep(), getNumberOfEvaluations(), init(), isMaximumNumberOfEvaluationsReached(), operator=(), optimize(), printPoint(), and bpp::DownhillSimplexMethod::tryExtrapolation().

◆ nbEvalMax_

unsigned int bpp::AbstractOptimizer::nbEvalMax_

protected

The maximum number of function evaluations allowed.

Definition at line 130 of file AbstractOptimizer.h.

Referenced by bpp::SimpleMultiDimensions::doInit(), bpp::SimpleNewtonMultiDimensions::doInit(), bpp::DownhillSimplexMethod::DownhillSimplexMethod(), bpp::GoldenSectionSearch::GoldenSectionSearch(), isMaximumNumberOfEvaluationsReached(), bpp::NewtonOneDimension::NewtonOneDimension(), operator=(), optimize(), and setMaximumNumberOfEvaluations().

◆ parameters_

ParameterList bpp::AbstractOptimizer::parameters_

private

The parameters that will be optimized.

Definition at line 72 of file AbstractOptimizer.h.

Referenced by autoParameter(), getParameter_(), getParameters(), getParameters_(), getParameterValue(), ignoreConstraints(), init(), operator=(), and step().

◆ profiler_

OutputStream* bpp::AbstractOptimizer::profiler_

private

The profiler.

Definition at line 82 of file AbstractOptimizer.h.

Referenced by getProfiler(), operator=(), profile(), profileln(), and setProfiler().

◆ startTime_

time_t bpp::AbstractOptimizer::startTime_

private

Definition at line 118 of file AbstractOptimizer.h.

Referenced by init(), operator=(), and printPoint().

◆ stepChar_

std::string bpp::AbstractOptimizer::stepChar_

private

Definition at line 124 of file AbstractOptimizer.h.

Referenced by getOptimizationProgressCharacter(), operator=(), and setOptimizationProgressCharacter().

◆ stopCondition_

OptimizationStopCondition* bpp::AbstractOptimizer::stopCondition_

private

The stoping condition to use while optimizing.

Definition at line 99 of file AbstractOptimizer.h.

Referenced by AbstractOptimizer(), getStopCondition(), init(), operator=(), setFunction(), setStopCondition(), step(), and ~AbstractOptimizer().

◆ tolIsReached_

bool bpp::AbstractOptimizer::tolIsReached_

protected

Tell if the tolerance level has been reached.

This field is initilaised by the init() method, maintained by the step() method and used in the optimize() method.

Definition at line 148 of file AbstractOptimizer.h.

Referenced by bpp::BfgsMultiDimensions::doStep(), bpp::ConjugateGradientMultiDimensions::doStep(), bpp::GoldenSectionSearch::doStep(), bpp::MetaOptimizer::doStep(), bpp::NewtonBacktrackOneDimension::doStep(), bpp::NewtonOneDimension::doStep(), bpp::SimpleMultiDimensions::doStep(), bpp::SimpleNewtonMultiDimensions::doStep(), init(), isToleranceReached(), operator=(), optimize(), and step().

◆ updateParameters_

bool bpp::AbstractOptimizer::updateParameters_

private

Definition at line 122 of file AbstractOptimizer.h.

Referenced by operator=(), step(), and updateParameters().

◆ verbose_

unsigned int bpp::AbstractOptimizer::verbose_

private

State of the verbose mode: > 0 = enabled.

This may not be used by the Optimizer.

Definition at line 111 of file AbstractOptimizer.h.

Referenced by getVerbose(), operator=(), optimize(), and setVerbose().

The documentation for this class was generated from the following files:

Bpp/Numeric/Function/AbstractOptimizer.h
Bpp/Numeric/Function/AbstractOptimizer.cpp

Public Member Functions

Protected Member Functions

Protected Attributes

Private Attributes

Detailed Description

Constructor & Destructor Documentation

◆ AbstractOptimizer() [1/2]

◆ AbstractOptimizer() [2/2]

◆ ~AbstractOptimizer()

Member Function Documentation

◆ addOptimizationListener()

◆ autoParameter()

◆ clone()

◆ doInit()

◆ doStep()

◆ fireOptimizationInitializationPerformed()

◆ fireOptimizationStepPerformed()

◆ getConstraintPolicy()

◆ getDefaultStopCondition() [1/2]

◆ getDefaultStopCondition() [2/2]

◆ getFunction() [1/2]

◆ getFunction() [2/2]

◆ getFunction_()

◆ getFunctionValue()

◆ getMessageHandler()

◆ getNumberOfEvaluations()

◆ getOptimizationProgressCharacter()

◆ getParameter_()

◆ getParameters()

◆ getParameters_()

◆ getParameterValue()

◆ getProfiler()

◆ getStopCondition() [1/2]

◆ getStopCondition() [2/2]

◆ getVerbose()

◆ hasFunction()

◆ ignoreConstraints()

◆ init()

◆ isInitialized()

◆ isMaximumNumberOfEvaluationsReached()

◆ isToleranceReached()

◆ listenerModifiesParameters()

◆ operator=()

◆ optimize()

◆ printMessage()

◆ printPoint()

◆ profile() [1/3]

◆ profile() [2/3]

◆ profile() [3/3]

◆ profileln() [1/3]

◆ profileln() [2/3]

◆ profileln() [3/3]

◆ setConstraintPolicy()

◆ setDefaultStopCondition_()

◆ setFunction()

◆ setMaximumNumberOfEvaluations()

◆ setMessageHandler()

◆ setOptimizationProgressCharacter()

◆ setProfiler()

◆ setStopCondition()

◆ setVerbose()

◆ step()

◆ updateParameters() [1/2]

◆ updateParameters() [2/2]

Member Data Documentation

◆ constraintPolicy_

◆ currentValue_

◆ defaultStopCondition_

◆ function_

◆ isInitialized_

◆ listeners_

◆ messageHandler_

◆ nbEval_

◆ nbEvalMax_

◆ parameters_

◆ profiler_

◆ startTime_

◆ stepChar_

◆ stopCondition_

◆ tolIsReached_