Fitness functions#

namespace gapp

class FitnessFunction#

#include <core/fitness_function.hpp>

template<typename T, size_t... ChromLens> class FitnessFunction : public gapp::FitnessFunctionBase<T>#

The base class of the fitness functions used in the GAs. The fitness functions take a candidate solution as an argument and return a fitness vector after evaluating the chromosomes of the candidate.

This class can only be used as the base class for fitness functions if their chromosome lengths are known at compile time. If the chromosome lengths are not known at compile time, use FitnessFunctionBase as the base class instead.

Template Parameters:

T – The gene type of the candidates expected by the fitness function.
ChromLens – The size of each chromosome of the candidate solutions expected by the fitness function. Each of the chromosome lengths must be at least 1. The number of chromosome lengths must be at least 1, but specifying more than 1 chromosome length is only allowed if T is a MixedGene type, and they must match the number of genes in the mixed gene type. For variable chromosome length problems, these values will only be used during the generation of the initial population.

Public Functions

inline constexpr FitnessFunction(Type type = Type::Static) noexcept#

Create a fitness function.

Parameters:: type – The type of the fitness function. The value should be either Type::Static or Type::Dynamic, based on whether the fitness function always returns the same fitness vector for a solution (static) or not (dynamic).

class FitnessFunctionBase#

#include <core/fitness_function.hpp>

template<typename T> class FitnessFunctionBase : public gapp::FitnessFunctionInfo #

The base class of the fitness functions used in the GAs. The fitness functions take a candidate solution as a parameter and return a fitness vector after evaluating the chromosomes of the candidate.

This class should be used as the base class for fitness functions if the chromosome lengths are not known at compile time. If the chromosome lengths are known at compile, use FitnessFunction as the base class instead.

Template Parameters:: T – The gene type of the candidates expected by the fitness function.

Subclassed by gapp::problems::BenchmarkFunction< IntegerGene >, gapp::problems::BenchmarkFunction< PermutationGene >, gapp::problems::BenchmarkFunction< BinaryGene >, gapp::FitnessFunction< T, ChromLens >, gapp::detail::FitnessLambda< T >, gapp::problems::BenchmarkFunction< T >

Public Types

using GeneType = T #: The gene type of the candidates that can be evaluated by the fitness function.

Public Functions

inline constexpr FitnessFunctionBase( Positive<size_t> chrom_len, Type type = Type::Static ) noexcept#

Create a fitness function.

Parameters:

chrom_len – The chromosome length expected by the fitness function, and which will be used for the candidate solutions in the GA. Must be at least 1, and a value must be specified even if the chromosome length is variable, as the value will still be used to generate the initial population.
type – The type of the fitness function. The value should be either Type::Static or Type::Dynamic, based on whether the fitness function always returns the same fitness vector for a solution (static) or not (dynamic).

inline constexpr FitnessFunctionBase( std::array<size_t, Candidate<T>::NumChroms> chrom_lens, Type type = Type::Static )#

Create a fitness function.

Parameters:

chrom_lens – The size of each of the chromosomes expected by the fitness function, and which will be used for the candidate solutions in the GA. Each chromosome length must be at least 1, and a value must be specified for each chromosome encoding the candidates, even if the chromosome length is variable, as the value will still be used to generate the initial population.
type – The type of the fitness function. The value should be either Type::Static or Type::Dynamic, based on whether the fitness function always returns the same fitness vector for a solution (static) or not (dynamic).

template<typename GeneType = T> constexpr size_t chrom_len() const noexcept#

Returns:: The chromosome length used by the fitness function.

template<typename GeneType> constexpr size_t chrom_len() const noexcept

Returns:: The length of the chromosome associated with the gene type GeneType.

inline FitnessVector operator()(const Candidate<T> &sol) const#

Compute the fitness value of a solution.

The length of the chromosomes will be equal to the chromosome lengths set for the fitness function, unless variable chromosome lengths are allowed.

Parameters:: sol – The candidate solution to evaluate.
Returns:: The fitness vector of the candidate, with a size equal to the number of objectives.

Private Functions

virtual FitnessVector invoke(const Candidate<T> &sol) const = 0#: The implementation of the fitness function. Should be thread-safe.

class FitnessFunctionInfo#

#include <core/fitness_function.hpp>

class FitnessFunctionInfo#

The base class of the fitness functions used in the GAs. Contains all of the properties of a fitness function that are not dependent on the gene type.

Fitness function implementations should not be derived directly from this class, but instead from FitnessFunctionBase or FitnessFunction.

Subclassed by gapp::FitnessFunctionBase< RealGene >, gapp::FitnessFunctionBase< BinaryGene >, gapp::FitnessFunctionBase< IntegerGene >, gapp::FitnessFunctionBase< PermutationGene >, gapp::FitnessFunctionBase< T >

Public Functions

inline constexpr const small_vector<size_t> &chrom_lens() const noexcept#

Returns:: The chromosome lengths used by the fitness function for each chromosome of the encoding.

inline constexpr bool is_dynamic() const noexcept#

Returns:: true if the fitness function is dynamic.

virtual ~FitnessFunctionInfo() = default#: Destructor.

Fitness functions

Contents

Fitness functions#

class FitnessFunction#

class FitnessFunctionBase#

class FitnessFunctionInfo#