grail::evolution::Mutation Struct Reference

The Mutation structure that encapsulates the mutation operation and its configuration in evolutionary algorithm. More...

#include <Mutation.hh>

Public Types
enum	MutationSelectionType { MutationSelectionType::ITERATE_AND_TEST_P, MutationSelectionType::SELECT_P_OF_N }
	Predefined ways in which individuals are selected from the population for mutation. Please refer to enum values for details about each of them. More...

Public Member Functions
	Mutation (double mutationPhaseRate=1.0, double mutateIndividualRate=0.09, double mutateGeneRate=0.5, bool mutateBothParentsAndChildren=true, MutationSelectionType randomResolutionTypeForIndividual=MutationSelectionType::SELECT_P_OF_N, MutationSelectionType randomResolutionTypeForGene=MutationSelectionType::SELECT_P_OF_N, std::mt19937_64::result_type seed=std::random_device{}())
	Mutation - Constructor. More...
virtual void	Perform (const std::vector< std::unique_ptr< Individual >> &parentPopulation, const std::vector< std::unique_ptr< Individual >> &childrenPopulation)
	Perform - Attempts to mutate the entire population. The umber of individuals mutates is based on mutation phase rate and mutate individual rate. More...

Public Attributes
double	MutationPhaseRate = 1.0
	Probability of a mutation phase to happen (globally). If a mutation is included in the algorithm, this value should most likely be equal to 1.0.
double	MutateIndividualRate = 0.09
	Probability that an individual will be considered for mutation [if (randomResolutionTypeForIndividual == ITERATE_AND_TEST_P)]; or the portion of individuals chosen for mutation [if (randomResolutionTypeForIndividual == SELECT_P_OF_N)].
double	MutateGeneRate = 0.5
	Probability that a gene of an individual will be considered for mutation [if (randomResolutionTypeForGene == ITERATE_AND_TEST_P)]; or the portion of genes within an individual chosen for mutation [if (randomResolutionTypeForGene == SELECT_P_OF_N)].
bool	MutateBothParentsAndChildren = false
	If true: both the previous population and children can be mutated; if false: only the individuals after crossover can be mutated.
MutationSelectionType	RandomResolutionTypeForIndividual = MutationSelectionType::SELECT_P_OF_N
	Specifies how individuals for mutations are chosen. See MutationSelectionType enum.
MutationSelectionType	RandomResolutionTypeForGene = MutationSelectionType::SELECT_P_OF_N
	Specifies how genes in an individual (already chosen for mutation) are chosen for mutation. See MutationSelectionType enum.

Protected Member Functions
virtual void	MutateIndividual (Individual &individual)
	The function that performs the mutation operation of one individual. More...
virtual void	MutateGene (Individual &individual, BaseEvoParam &gene)
	The function that performs the mutation operation of a single parameter within the mutated individual. More...

Protected Attributes
std::mt19937_64	randGen

Friends
class	EAOptimizer

Detailed Description

The Mutation structure that encapsulates the mutation operation and its configuration in evolutionary algorithm.

Member Enumeration Documentation

MutationSelectionType

enum grail::evolution::Mutation::MutationSelectionType

strong

Predefined ways in which individuals are selected from the population for mutation. Please refer to enum values for details about each of them.

Enumerator
ITERATE_AND_TEST_P	ITERATE_AND_TEST_P - a random test whether to mutate an individual will be performed for each individual in a loop.
SELECT_P_OF_N	SELECT_P_OF_N - upfront, the PopulationSize*MutationIndividualRate of individuals will be randomly chosen for mutation. This ensures that the expected number of individuals will be mutated, regardless of RNG.

Constructor & Destructor Documentation

Mutation()

grail::evolution::Mutation::Mutation	(	double	mutationPhaseRate = 1.0,
		double	mutateIndividualRate = 0.09,
		double	mutateGeneRate = 0.5,
		bool	mutateBothParentsAndChildren = true,
		MutationSelectionType	randomResolutionTypeForIndividual = MutationSelectionType::SELECT_P_OF_N,
		MutationSelectionType	randomResolutionTypeForGene = MutationSelectionType::SELECT_P_OF_N,
		std::mt19937_64::result_type	seed = std::random_device{}()
	)

Mutation - Constructor.

Parameters

mutationPhaseRate	- the probability of a mutation phase to happen (globally). If a mutation is included in the algorithm, this value should most likely be equal to 1.0.
mutateIndividualRate	- the probability that an individual will be considered for mutation [if (randomResolutionTypeForIndividual == ITERATE_AND_TEST_P)]; or represents the portion of individuals chosen for mutation [if (randomResolutionTypeForIndividual == SELECT_P_OF_N)].
mutateGeneRate	- the probability that a gene of an individual will be considered for mutation [if (randomResolutionTypeForGene == ITERATE_AND_TEST_P)]; or the portion of genes within an individual chosen for mutation [if (randomResolutionTypeForGene == SELECT_P_OF_N)].
mutateBothParentsAndChildren	- if true: both the previous population and children can be mutated; if false: only the individuals after crossover can be mutated.
randomResolutionTypeForIndividual	- specifies how individuals for mutations are chosen. See MutationSelectionType enum.
randomResolutionTypeForGene	- specifies how genes in an individual (already chosen for mutation) are chosen for mutation. See MutationSelectionType enum.
seed	- the initial seed for random numbers generator.

Member Function Documentation

MutateGene()

void grail::evolution::Mutation::MutateGene ( Individual &  individual, BaseEvoParam &  gene  )

protectedvirtual

The function that performs the mutation operation of a single parameter within the mutated individual.

Parameters

individual	- the individual to be mutated.
gene	- the parameter of the individual to be mutated.

MutateIndividual()

void grail::evolution::Mutation::MutateIndividual ( Individual &  individual )

protectedvirtual

The function that performs the mutation operation of one individual.

Parameters

individual

- the individual to be mutated.

Perform()

void grail::evolution::Mutation::Perform ( const std::vector< std::unique_ptr< Individual >> &  parentPopulation, const std::vector< std::unique_ptr< Individual >> &  childrenPopulation  )

virtual

Perform - Attempts to mutate the entire population. The umber of individuals mutates is based on mutation phase rate and mutate individual rate.

Parameters

parentPopulation	Parent population - before the mutation.
childrenPopulation	Children population - after the mutation.

---

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

• GrailEvolution/Mutation.hh
• GrailEvolution/Mutation.cpp