genetiK::EvolutionaryAlgorithm Class Reference
[Common GenetiK framework]

EvolutionaryAlgorithm: The main evolutionary algorithm. This class represents the main evolutionary algorithm. More...

#include <EvolutionaryAlgorithm.h>

Collaboration diagram for genetiK::EvolutionaryAlgorithm:

[legend]List of all members.

Public Member Functions
	EvolutionaryAlgorithm (const unsigned int populationSize, IndividualFactory *factory, StopCriterion *stopCriterion=NULL, SelectionMethod *selectionMethod=NULL, const double mutationProbability=0.01, const double crossOverProbability=0.7, const bool elitism=true)
virtual int	run ()
virtual int	generatePopulation ()
virtual int	evolve ()
virtual Population *	getPopulation () const
Protected Attributes
StopCriterion *	stopCriterion
	Specifies a criterion to stop the EvolutionaryAlgorithm.
Population *	population
	Contains the individuals generated by this Algorithm at the present generation.
IndividualFactory *	individualFactory
	The Factory that is used to create specific Invidual objects.
SelectionMethod *	selectionMethod
	Provides a scheme to select individuals for crossover and reproduction.
unsigned int	populationSize
	The size of the Population of Individuals at each generation.
double	crossOverProbabilty
	Probability to perform crossover for a couple of Individuals.
double	mutationProbabilty
	Probability to perform mutation on an Individual.
bool	elitism
	Specifies if the current EvolutionaryAlgorithm should be elitist.

---

Detailed Description

EvolutionaryAlgorithm: The main evolutionary algorithm. This class represents the main evolutionary algorithm.

---

Constructor & Destructor Documentation

genetiK::EvolutionaryAlgorithm::EvolutionaryAlgorithm (  const unsigned int  populationSize, IndividualFactory *  factory, StopCriterion *  stopCriterion = NULL, SelectionMethod *  selectionMethod = NULL, const double  mutationProbability = 0.01, const double  crossOverProbability = 0.7, const bool  elitism = true )

Parameters: populationSize  size of the population factory  the individual factory, redefine yours to generate your own Individual. stopCriterion  this class decides the end of the algortihm. selectionMethod  the selection method, if you don't specify it, the default RouletteWheel will be used. The genetiK library has genetiK::RouletteWheel, genetiK::TournamentSelection and RankingSelection. mutationProbability  0.01 by default. crossOverProbability  0.7 by default. elitism  if this is true, the best individual of the current generation will be copied to the next one. See also: Version

---

Member Function Documentation

int genetiK::EvolutionaryAlgorithm::evolve (   )  [virtual]

Usually this method is called by .run(), but if you want more control on the evolution, you can manually call this after generatePopulation() Exceptions: IllegalStateException  An IllegalStateException is thrown if a call to evolve() is made without having previously called generatePopulation() Returns: return 0 if it's all ok.

int genetiK::EvolutionaryAlgorithm::generatePopulation (   )  [virtual]

Usually this method is called by .run(), but if you want more control on the evolution, you can manually call this before evolve() Returns: return 0 if it's all ok.

virtual Population* genetiK::EvolutionaryAlgorithm::getPopulation (   )  const [inline, virtual]

This method is used to access the population at current generation Returns: return the population itself.

int genetiK::EvolutionaryAlgorithm::run (   )  [virtual]

Call this method to start the algorithm. Returns: return 0 if it's all ok.

---

Member Data Documentation

bool genetiK::EvolutionaryAlgorithm::elitism [protected]

Specifies if the current EvolutionaryAlgorithm should be elitist. If elitism is applied, the best Individual in each generation alway survives, unmodified, in the next one.

IndividualFactory* genetiK::EvolutionaryAlgorithm::individualFactory [protected]

The Factory that is used to create specific Invidual objects. This class should be user supplied to create the desired type of Inviduals.

---

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

• devel/genetik/src/EvolutionaryAlgorithm.h
• devel/genetik/src/EvolutionaryAlgorithm.cpp

---