GeneralGA Class Reference
[Genetic algorithm optimization]

General genetic algorithm class. More...

#include <GeneralGA.h>

Inheritance diagram for GeneralGA:

List of all members.

Public Types

typedef std::vector< gplib::rvec > tparamvector

We provide some typedefs that are used in other parts as well.

typedef std::vector
< boost::shared_ptr
< GeneralObjective > > tObjectiveVector

typedef std::vector
< std::vector< int > > tparamindv

Public Member Functions

const std::vector< double > & GetWeights ()

Return the weight for each objecive function.

void SetupParams (const ttranscribed &member, tparamvector &params)

Copy the appropriate parameter values from the transcribed vector for use with the objective functions.

const std::vector< double > & GetAvgFit ()

Return the average fitness for each objective function.

void SetThreads (const int LocalThreads)

Set the number of threads, only has an effect if openmp is enabled.

void SetElitist (const bool IsElitist)

Do we want elitist behaviour, effect depends on the GA implementetion in the derived class.

void SetWeights (const std::vector< double > &LocalWeights)

Set the weights for each objective function.

void SetParameterIndices (const tparamindv &Indices)

Configure which parts of the complete parameter vector are used in each objective function.

void PrintFitStat (std::ostream &output)

Print Fitness statistics for each objective function to output.

void PrintMisfit (std::ostream &output)

Print misfit for each member to output.

void PrintTranscribed (std::ostream &output)

Print transcribed values for each member to output.

void PrintUniquePop (std::ostream &output)

Print each population member exactly once.

void PrintBestMisfit (std::ostream &output)

Print misfit of the best population members.

unsigned virtual int GetNBestmodels ()=0

This has to be implemented in the derived class to return the number of best models.

virtual std::vector< int > GetBestModelIndices ()=0

Return the indices of the best models.

virtual void DoIteration (const int iterationnumber, const bool last)

Do one iteration of the GA.

virtual void CalcProbabilities (const int iterationnumber, gplib::rmat &LocalMisFit, GeneralPopulation &LocalPopulation)=0

Calculate the Probabilities.

GeneralGA (GeneralPropagation *const LocalPropagation, GeneralPopulation *const LocalPopulation, GeneralTranscribe *const LocalTranscribe, const tObjectiveVector &IndObjective)

virtual ~GeneralGA ()

Protected Member Functions

virtual void Elitism (const int iterationnumber)

default implementation does nothing, this can be overriden

Protected Attributes

gplib::rmat OldMisFit

const unsigned int nobjective

The number of objective functions we're using.

gplib::rmat Transcribed

A matrix holding the numerical model vectors (transcribed from the genes) for each population member.

gplib::rmat MisFit

Misfit first index objective function second index population member.

std::vector< tObjectiveVector > Objective

Each population member gets its own instance of objective function objects for parallel calculations.

GeneralPropagation *const Propagation

A pointer to an object dealing with propagation.

GeneralPopulation *const Population

A pointer to an object holding the population.

GeneralTranscribe *const Transcribe

A pointer to an object translating genes to model vectors.

Detailed Description

General genetic algorithm class.

GeneralGA implements the functionality that is common to all genetic algorithms it declares a bunch of virtual functions, that are implemented in derived classes and pointers to several other objects that together make up the genetic algorithm

Definition at line 25 of file GeneralGA.h.

Member Typedef Documentation

typedef std::vector<gplib::rvec > GeneralGA::tparamvector

We provide some typedefs that are used in other parts as well.

Definition at line 28 of file GeneralGA.h.

typedef std::vector<boost::shared_ptr<GeneralObjective> > GeneralGA::tObjectiveVector

Definition at line 29 of file GeneralGA.h.

typedef std::vector<std::vector<int> > GeneralGA::tparamindv

Definition at line 30 of file GeneralGA.h.

Constructor & Destructor Documentation

GeneralGA::GeneralGA	(	GeneralPropagation *const	LocalPropagation,
		GeneralPopulation *const	LocalPopulation,
		GeneralTranscribe *const	LocalTranscribe,
		const tObjectiveVector &	IndObjective
	)

Generate a separate objective function object for each population member, simplifies parallelization

Definition at line 37 of file GeneralGA.cpp.

References GeneralPopulation::GetPopsize(), and Objective.

GeneralGA::~GeneralGA ( ) [virtual]

Definition at line 59 of file GeneralGA.cpp.

Member Function Documentation

virtual void GeneralGA::Elitism ( const int iterationnumber ) [inline, protected, virtual]

default implementation does nothing, this can be overriden

See also:: ParetoGA

Reimplemented in AnnealingGA, and ParetoGA.

Definition at line 74 of file GeneralGA.h.

Referenced by DoIteration().

const std::vector<double>& GeneralGA::GetWeights ( ) [inline]

Return the weight for each objecive function.

Definition at line 77 of file GeneralGA.h.

void GeneralGA::SetupParams	(	const ttranscribed &	member,
		tparamvector &	params
	)

Copy the appropriate parameter values from the transcribed vector for use with the objective functions.

Definition at line 63 of file GeneralGA.cpp.

References nobjective.

const std::vector<double>& GeneralGA::GetAvgFit ( ) [inline]

Return the average fitness for each objective function.

Definition at line 81 of file GeneralGA.h.

void GeneralGA::SetThreads ( const int LocalThreads ) [inline]

Set the number of threads, only has an effect if openmp is enabled.

Definition at line 83 of file GeneralGA.h.

void GeneralGA::SetElitist ( const bool IsElitist ) [inline]

Do we want elitist behaviour, effect depends on the GA implementetion in the derived class.

Definition at line 85 of file GeneralGA.h.

Referenced by main().

void GeneralGA::SetWeights ( const std::vector< double > & LocalWeights )

Set the weights for each objective function.

Definition at line 72 of file GeneralGA.cpp.

void GeneralGA::SetParameterIndices ( const tparamindv & Indices )

Configure which parts of the complete parameter vector are used in each objective function.

Definition at line 79 of file GeneralGA.cpp.

Referenced by main().

void GeneralGA::PrintFitStat ( std::ostream & output )

Print Fitness statistics for each objective function to output.

Definition at line 84 of file GeneralGA.cpp.

References nobjective.

void GeneralGA::PrintMisfit ( std::ostream & output )

Print misfit for each member to output.

Definition at line 94 of file GeneralGA.cpp.

References MisFit.

Referenced by main().

void GeneralGA::PrintTranscribed ( std::ostream & output )

Print transcribed values for each member to output.

Definition at line 109 of file GeneralGA.cpp.

References Transcribed.

void GeneralGA::PrintUniquePop ( std::ostream & output ) [inline]

Print each population member exactly once.

Definition at line 97 of file GeneralGA.h.

References UniquePop::PrintAll().

void GeneralGA::PrintBestMisfit ( std::ostream & output )

Print misfit of the best population members.

Definition at line 123 of file GeneralGA.cpp.

References GetBestModelIndices(), and MisFit.

unsigned virtual int GeneralGA::GetNBestmodels ( ) [pure virtual]

This has to be implemented in the derived class to return the number of best models.

Implemented in AnnealingGA, and ParetoGA.

virtual std::vector<int> GeneralGA::GetBestModelIndices ( ) [pure virtual]

Return the indices of the best models.

Implemented in AnnealingGA, and ParetoGA.

Referenced by PrintBestMisfit().

void GeneralGA::DoIteration	(	const int	iterationnumber,
		const bool	last
	)			`[virtual]`

Do one iteration of the GA.

Definition at line 237 of file GeneralGA.cpp.

References CalcProbabilities(), Elitism(), Mean(), MisFit, GeneralPropagation::NextGeneration(), OldMisFit, Population, Propagation, and GeneralPopulation::StoreOldPopulation().

Referenced by main().

virtual void GeneralGA::CalcProbabilities	(	const int	iterationnumber,
		gplib::rmat &	LocalMisFit,
		GeneralPopulation &	LocalPopulation
	)			`[pure virtual]`

Calculate the Probabilities.

Implemented in AnnealingGA, and ParetoGA.

Referenced by DoIteration().

Member Data Documentation

gplib::rmat GeneralGA::OldMisFit [protected]

Definition at line 58 of file GeneralGA.h.

Referenced by DoIteration(), and ParetoGA::Elitism().

const unsigned int GeneralGA::nobjective [protected]

The number of objective functions we're using.

Definition at line 60 of file GeneralGA.h.

Referenced by PrintFitStat(), and SetupParams().

gplib::rmat GeneralGA::Transcribed [protected]

A matrix holding the numerical model vectors (transcribed from the genes) for each population member.

Definition at line 62 of file GeneralGA.h.

Referenced by PrintTranscribed().

gplib::rmat GeneralGA::MisFit [protected]

Misfit first index objective function second index population member.

Definition at line 64 of file GeneralGA.h.

Referenced by DoIteration(), ParetoGA::Elitism(), PrintBestMisfit(), ParetoGA::PrintFront(), PrintMisfit(), and ParetoGA::PrintRanks().

std::vector< tObjectiveVector > GeneralGA::Objective [protected]

Each population member gets its own instance of objective function objects for parallel calculations.

Definition at line 66 of file GeneralGA.h.

Referenced by GeneralGA().

GeneralPropagation* const GeneralGA::Propagation [protected]

A pointer to an object dealing with propagation.

Definition at line 68 of file GeneralGA.h.

Referenced by DoIteration().

GeneralPopulation* const GeneralGA::Population [protected]

A pointer to an object holding the population.

Definition at line 70 of file GeneralGA.h.

Referenced by DoIteration(), ParetoGA::Elitism(), and AnnealingGA::GetBestIndex().

GeneralTranscribe* const GeneralGA::Transcribe [protected]

A pointer to an object translating genes to model vectors.

Definition at line 72 of file GeneralGA.h.

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


Public Types
typedef std::vector< gplib::rvec >	tparamvector
	We provide some typedefs that are used in other parts as well.
typedef std::vector < boost::shared_ptr < GeneralObjective > >	tObjectiveVector
typedef std::vector < std::vector< int > >	tparamindv
Public Member Functions
const std::vector< double > &	GetWeights ()
	Return the weight for each objecive function.
void	SetupParams (const ttranscribed &member, tparamvector &params)
	Copy the appropriate parameter values from the transcribed vector for use with the objective functions.
const std::vector< double > &	GetAvgFit ()
	Return the average fitness for each objective function.
void	SetThreads (const int LocalThreads)
	Set the number of threads, only has an effect if openmp is enabled.
void	SetElitist (const bool IsElitist)
	Do we want elitist behaviour, effect depends on the GA implementetion in the derived class.
void	SetWeights (const std::vector< double > &LocalWeights)
	Set the weights for each objective function.
void	SetParameterIndices (const tparamindv &Indices)
	Configure which parts of the complete parameter vector are used in each objective function.
void	PrintFitStat (std::ostream &output)
	Print Fitness statistics for each objective function to output.
void	PrintMisfit (std::ostream &output)
	Print misfit for each member to output.
void	PrintTranscribed (std::ostream &output)
	Print transcribed values for each member to output.
void	PrintUniquePop (std::ostream &output)
	Print each population member exactly once.
void	PrintBestMisfit (std::ostream &output)
	Print misfit of the best population members.
unsigned virtual int	GetNBestmodels ()=0
	This has to be implemented in the derived class to return the number of best models.
virtual std::vector< int >	GetBestModelIndices ()=0
	Return the indices of the best models.
virtual void	DoIteration (const int iterationnumber, const bool last)
	Do one iteration of the GA.
virtual void	CalcProbabilities (const int iterationnumber, gplib::rmat &LocalMisFit, GeneralPopulation &LocalPopulation)=0
	Calculate the Probabilities.
	GeneralGA (GeneralPropagation const LocalPropagation, GeneralPopulation const LocalPopulation, GeneralTranscribe *const LocalTranscribe, const tObjectiveVector &IndObjective)
virtual	~GeneralGA ()
Protected Member Functions
virtual void	Elitism (const int iterationnumber)
	default implementation does nothing, this can be overriden
Protected Attributes
gplib::rmat	OldMisFit
const unsigned int	nobjective
	The number of objective functions we're using.
gplib::rmat	Transcribed
	A matrix holding the numerical model vectors (transcribed from the genes) for each population member.
gplib::rmat	MisFit
	Misfit first index objective function second index population member.
std::vector< tObjectiveVector >	Objective
	Each population member gets its own instance of objective function objects for parallel calculations.
GeneralPropagation *const	Propagation
	A pointer to an object dealing with propagation.
GeneralPopulation *const	Population
	A pointer to an object holding the population.
GeneralTranscribe *const	Transcribe
	A pointer to an object translating genes to model vectors.

GeneralGA Class Reference [Genetic algorithm optimization]

Public Types

Public Member Functions

Protected Member Functions

Protected Attributes

Detailed Description

Member Typedef Documentation

Constructor & Destructor Documentation

Member Function Documentation

Member Data Documentation

GeneralGA Class Reference
[Genetic algorithm optimization]