genetic_algorithm

package module
v0.0.0-...-0872a5e Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Dec 15, 2014 License: MIT Imports: 8 Imported by: 6

README

genetic_algorithm

Genetic algorithm library. Written in the Go programming language (Golang).

Installation

If you don't have the Go development environment installed, visit the Getting Started document and follow the instructions. Once you're ready, execute the following command:

go get -u github.com/WiseBird/genetic_algorithm

Documentation

Index

Constants

View Source 
const (
	// Will roll for every element in chromosome, mutate it if success
	MutatorOneByOneType = 0
	// Will mutete exactly (Npop * Nel * P) elements
	MutatorExactCountType = 1
)

Variables

View Source 
var (
	NopMutator = &nopMutator{}
)

Functions

func NewOrderCrossoverVer1 

func NewOrderCrossoverVer1() *orderCrossover

func NewOrderCrossoverVer2 

func NewOrderCrossoverVer2() *orderCrossover

Types

type BinaryChromosome 

type BinaryChromosome struct {
	*ChromosomeBase
	// contains filtered or unexported fields
}

func NewBinaryChromosome 

func NewBinaryChromosome(genes BinaryGenes) *BinaryChromosome

func (*BinaryChromosome) BinaryGenes 

func (chrom *BinaryChromosome) BinaryGenes() BinaryGenes

func (*BinaryChromosome) Genes 

func (chrom *BinaryChromosome) Genes() GenesInterface

func (*BinaryChromosome) String 

func (chrom *BinaryChromosome) String() string

type BinaryGenes 

type BinaryGenes []bool

func (BinaryGenes) Copy 

func (g BinaryGenes) Copy(genes GenesInterface, from1, from2, to2 int) int

func (BinaryGenes) Get 

func (g BinaryGenes) Get(i int) interface{}

func (BinaryGenes) Len 

func (g BinaryGenes) Len() int

func (BinaryGenes) Set 

func (g BinaryGenes) Set(i int, val interface{})

func (BinaryGenes) Swap 

func (g BinaryGenes) Swap(i, j int)

type BinaryMutator 

type BinaryMutator struct {
}

Mutator for binary chromosomes Simply inverts specified bit

func (*BinaryMutator) MutateCromosome 

func (mutator *BinaryMutator) MutateCromosome(chrom ChromosomeInterface, ind int)

type BinaryRandomInitializer 

type BinaryRandomInitializer struct {
}

func NewBinaryRandomInitializer 

func NewBinaryRandomInitializer() *BinaryRandomInitializer

func (*BinaryRandomInitializer) Init 

func (initializer *BinaryRandomInitializer) Init(count, chromSize int) Chromosomes

type ChromosomeBase 

type ChromosomeBase struct {
	// contains filtered or unexported fields
}

Base class for chromosomes.

func NewChromosomeBase 

func NewChromosomeBase() *ChromosomeBase

func (*ChromosomeBase) Cost 

func (chrom *ChromosomeBase) Cost() float64

func (*ChromosomeBase) SetCost 

func (chrom *ChromosomeBase) SetCost(cost float64)

type ChromosomeInterface 

type ChromosomeInterface interface {
	Genes() GenesInterface
	SetCost(float64)
	Cost() float64
}

func NewEmptyBinaryChromosome 

func NewEmptyBinaryChromosome(genesLen int) ChromosomeInterface

func NewEmptyOrderedChromosome 

func NewEmptyOrderedChromosome(genesLen int) ChromosomeInterface

type Chromosomes 

type Chromosomes []ChromosomeInterface

func (Chromosomes) Len 

func (c Chromosomes) Len() int

func (Chromosomes) Less 

func (c Chromosomes) Less(i, j int) bool

func (Chromosomes) MeanCost 

func (c Chromosomes) MeanCost() float64

func (Chromosomes) SetCost 

func (c Chromosomes) SetCost(cost CostFunction)

func (Chromosomes) String 

func (c Chromosomes) String() string

func (Chromosomes) Swap 

func (c Chromosomes) Swap(i, j int)

type CostFunction 

type CostFunction func(ChromosomeInterface) float64

type CrossoverInterface 

type CrossoverInterface interface {
	ParentsCount() int
	Crossover(Chromosomes) Chromosomes
}

type CycleCrossover 

type CycleCrossover struct {
}

Crossover for ordered chromosomes. Tends to preserve absolute order.

func NewCycleCrossover 

func NewCycleCrossover() *CycleCrossover

func (*CycleCrossover) Crossover 

func (crossover *CycleCrossover) Crossover(parents Chromosomes) Chromosomes

func (*CycleCrossover) ParentsCount 

func (crossover *CycleCrossover) ParentsCount() int

type DisplacementMutator 

type DisplacementMutator struct {
	*MutatorIntervalBase
}

Mutator selects some part of the chromosome and places it in random position

func NewDisplacementMutator 

func NewDisplacementMutator(probability float64, chromosomeConstructor EmptyChromosomeConstructor) *DisplacementMutator

Probability is applied to each chromosome

func (*DisplacementMutator) MutateGenes 

func (mutator *DisplacementMutator) MutateGenes(genes GenesInterface, from, to int)

type EdgeRecombinationCrossover 

type EdgeRecombinationCrossover struct {
}

The edge recombination operator (ERO) is an operator that creates a path that is similar to a set of existing paths (parents) by looking at the edges rather than the vertices.

http://en.wikipedia.org/wiki/Edge_recombination_operator

func NewEdgeRecombinationCrossover 

func NewEdgeRecombinationCrossover() *EdgeRecombinationCrossover

func (*EdgeRecombinationCrossover) Crossover 

func (crossover *EdgeRecombinationCrossover) Crossover(parents Chromosomes) Chromosomes

func (*EdgeRecombinationCrossover) ParentsCount 

func (crossover *EdgeRecombinationCrossover) ParentsCount() int

type EmptyChromosomeConstructor 

type EmptyChromosomeConstructor func(genesLen int) ChromosomeInterface

type GenesInterface 

type GenesInterface interface {
	Len() int
	Copy(genes GenesInterface, from1, from2, to2 int) int
	Swap(int, int)
	Get(int) interface{}
	Set(int, interface{})
}

type HierarchicalDuration 

type HierarchicalDuration struct {
	Name     string
	Duration time.Duration
	Calls    int

	Children map[string]*HierarchicalDuration
}

func (*HierarchicalDuration) String 

func (hierarchy *HierarchicalDuration) String() string

type IncrementalOptimizer 

type IncrementalOptimizer struct {
	*OptimizerBase
	// contains filtered or unexported fields
}

func NewIncrementalOptimizer 

func NewIncrementalOptimizer() *IncrementalOptimizer

func (*IncrementalOptimizer) Weeder 

func (optimizer *IncrementalOptimizer) Weeder(weeder WeederInterface) *IncrementalOptimizer

type InitializerInterface 

type InitializerInterface interface {
	Init(count, chromSize int) Chromosomes
}

type InvertDisplacementMutator 

type InvertDisplacementMutator struct {
	*MutatorIntervalBase
	// contains filtered or unexported fields
}

InvertMutator + DisplacementMutator Mutator selects some part of the chromosome inverts it then place at other position

func NewInvertDisplacementMutator 

func NewInvertDisplacementMutator(probability float64, chromosomeConstructor EmptyChromosomeConstructor) *InvertDisplacementMutator

Probability is applied to each chromosome

func (*InvertDisplacementMutator) MutateGenes 

func (mutator *InvertDisplacementMutator) MutateGenes(genes GenesInterface, from, to int)

type InvertMutator 

type InvertMutator struct {
	*MutatorIntervalBase
}

Mutator selects some part of the chromosome and inverts it

func NewInvertMutator 

func NewInvertMutator(probability float64, chromosomeConstructor EmptyChromosomeConstructor) *InvertMutator

Probability is applied to each chromosome

func (*InvertMutator) MutateGenes 

func (mutator *InvertMutator) MutateGenes(genes GenesInterface, from, to int)

type InvertSwapMutator 

type InvertSwapMutator struct {
	*MutatorIntervalBase
	// contains filtered or unexported fields
}

InvertMutator + SwapMutator Mutator selects some part of the chromosome inverts it then swap on gene from it with one gene outside of the interval

func NewInvertSwapMutator 

func NewInvertSwapMutator(probability float64, chromosomeConstructor EmptyChromosomeConstructor) *InvertSwapMutator

Probability is applied to each chromosome

func (*InvertSwapMutator) MutateGenes 

func (mutator *InvertSwapMutator) MutateGenes(genes GenesInterface, from, to int)

type MultiPointCrossover 

type MultiPointCrossover struct {
	// contains filtered or unexported fields
}

func NewMultiPointCrossover 

func NewMultiPointCrossover(chromConstr EmptyChromosomeConstructor, crossPointsCount int) *MultiPointCrossover

func NewOnePointCrossover 

func NewOnePointCrossover(chromConstr EmptyChromosomeConstructor) *MultiPointCrossover

func NewTwoPointCrossover 

func NewTwoPointCrossover(chromConstr EmptyChromosomeConstructor) *MultiPointCrossover

func (*MultiPointCrossover) CanProduceCopiesOfParents 

func (crossover *MultiPointCrossover) CanProduceCopiesOfParents(val bool) *MultiPointCrossover

func (*MultiPointCrossover) Crossover 

func (crossover *MultiPointCrossover) Crossover(parents Chromosomes) Chromosomes

func (*MultiPointCrossover) ParentsCount 

func (crossover *MultiPointCrossover) ParentsCount() int

type MutatorGeneBase 

type MutatorGeneBase struct {
	MutatorGeneBaseVirtualMInterface
	// contains filtered or unexported fields
}

Base class for mutators that mutate separate genes

func NewBinaryMutator 

func NewBinaryMutator(probability float64) *MutatorGeneBase

Probability is applied to each element separately.

func NewGeneBaseMutator 

func NewGeneBaseMutator(virtual MutatorGeneBaseVirtualMInterface, probability float64) *MutatorGeneBase

func NewSwapMutator 

func NewSwapMutator(probability float64) *MutatorGeneBase

Probability is applied to each element separately.

func (*MutatorGeneBase) ExactCount 

func (mutator *MutatorGeneBase) ExactCount() *MutatorGeneBase

Will mutete exactly (Npop * Nel * P) elements

func (*MutatorGeneBase) Mutate 

func (mutator *MutatorGeneBase) Mutate(population Chromosomes)

func (*MutatorGeneBase) OneByOne 

func (mutator *MutatorGeneBase) OneByOne() *MutatorGeneBase

Will roll for every element in chromosome, mutate it if success

func (*MutatorGeneBase) WithElitism 

func (mutator *MutatorGeneBase) WithElitism(count int) *MutatorGeneBase

The best chromosome[s] can't be mutated

func (*MutatorGeneBase) WithoutElitism 

func (mutator *MutatorGeneBase) WithoutElitism() *MutatorGeneBase

All chromosomes can be mutated

type MutatorGeneBaseVirtualMInterface 

type MutatorGeneBaseVirtualMInterface interface {
	MutateCromosome(chrom ChromosomeInterface, ind int)
}

MutatorGeneBase's virtual methods

type MutatorInterface 

type MutatorInterface interface {
	Mutate(Chromosomes)
}

Mutator interface

type MutatorIntervalBase 

type MutatorIntervalBase struct {
	MutatorIntervalBaseVirtualMInterface
	// contains filtered or unexported fields
}

Base class for mutators that mutate some interval of genes

func NewInsertionMutator 

func NewInsertionMutator(probability float64, chromosomeConstructor EmptyChromosomeConstructor) *MutatorIntervalBase

Mutator selects random element from chromosome and places it in random position Probability is applied to each chromosome

func NewMutatorIntervalBase 

func NewMutatorIntervalBase(virtual MutatorIntervalBaseVirtualMInterface, probability float64, chromosomeConstructor EmptyChromosomeConstructor) *MutatorIntervalBase

Probability is applied to each chromosome By default interval will be equal one third of the chromosome

func (*MutatorIntervalBase) ExactInterval 

func (mutator *MutatorIntervalBase) ExactInterval(from, to int) *MutatorIntervalBase

Sets interval len to be randomly choosen from [from:to]

func (*MutatorIntervalBase) Mutate 

func (mutator *MutatorIntervalBase) Mutate(population Chromosomes)

func (*MutatorIntervalBase) PercentageInterval 

func (mutator *MutatorIntervalBase) PercentageInterval(from, to float64) *MutatorIntervalBase

Sets interval len to be randomly choosen from [from*chromLen : to*chromLen]

type MutatorIntervalBaseVirtualMInterface 

type MutatorIntervalBaseVirtualMInterface interface {
	MutateGenes(genes GenesInterface, from, to int)
}

MutatorIntervalBase's virtual methods

type OptimizerAggregator 

type OptimizerAggregator struct {
	// contains filtered or unexported fields
}

func NewOptimizerAggregator 

func NewOptimizerAggregator() *OptimizerAggregator

func (*OptimizerAggregator) Iterations 

func (aggregator *OptimizerAggregator) Iterations(iterations int) *OptimizerAggregator

func (*OptimizerAggregator) Optimize 

func (aggregator *OptimizerAggregator) Optimize() (ChromosomeInterface, StatisticsDataInterface)

func (*OptimizerAggregator) Optimizer 

func (aggregator *OptimizerAggregator) Optimizer(optimizer OptimizerInterface) *OptimizerAggregator

func (*OptimizerAggregator) StatisticsAggregatorConstructor 

func (aggregator *OptimizerAggregator) StatisticsAggregatorConstructor(constr StatisticsAggregatorConstructor) *OptimizerAggregator

func (*OptimizerAggregator) StatisticsOptions 

func (aggregator *OptimizerAggregator) StatisticsOptions(statisticsOptions StatisticsOptionsInterface) *OptimizerAggregator

type OptimizerBase 

type OptimizerBase struct {
	OptimizerBaseVirtualMInterface
	// contains filtered or unexported fields
}

func NewOptimizerBase 

func NewOptimizerBase(virtual OptimizerBaseVirtualMInterface) *OptimizerBase

func (*OptimizerBase) ChromSize 

func (optimizer *OptimizerBase) ChromSize(chromSize int) *OptimizerBase

func (*OptimizerBase) CostFunction 

func (optimizer *OptimizerBase) CostFunction(cost CostFunction) *OptimizerBase

func (*OptimizerBase) Crossover 

func (optimizer *OptimizerBase) Crossover(crossover CrossoverInterface) *OptimizerBase

func (*OptimizerBase) Initializer 

func (optimizer *OptimizerBase) Initializer(initializer InitializerInterface) *OptimizerBase

func (*OptimizerBase) Mutator 

func (optimizer *OptimizerBase) Mutator(mutator MutatorInterface) *OptimizerBase

func (*OptimizerBase) Optimize 

func (optimizer *OptimizerBase) Optimize() (ChromosomeInterface, StatisticsDataInterface)

func (*OptimizerBase) PopSize 

func (optimizer *OptimizerBase) PopSize(popSize int) *OptimizerBase

func (*OptimizerBase) Selector 

func (optimizer *OptimizerBase) Selector(selector SelectorInterface) *OptimizerBase

func (*OptimizerBase) SetupStatisticsOptions 

func (optimizer *OptimizerBase) SetupStatisticsOptions() StatisticsOptionsInterface

func (*OptimizerBase) StatisticsConstructor 

func (optimizer *OptimizerBase) StatisticsConstructor(constr StatisticsConstructor) *OptimizerBase

func (*OptimizerBase) StatisticsOptions 

func (optimizer *OptimizerBase) StatisticsOptions(statisticsOptions StatisticsOptionsInterface) *OptimizerBase

func (*OptimizerBase) StopCriterion 

func (optimizer *OptimizerBase) StopCriterion(stopCriterion StopCriterionInterface) *OptimizerBase

type OptimizerBaseVirtualMInterface 

type OptimizerBaseVirtualMInterface interface {
	// contains filtered or unexported methods
}

MutatorBase's virtual methods

type OptimizerInterface 

type OptimizerInterface interface {
	Optimize() (ChromosomeInterface, StatisticsDataInterface)
}

type OptimizerWithStatisticsOptionsSetup 

type OptimizerWithStatisticsOptionsSetup interface {
	SetupStatisticsOptions() StatisticsOptionsInterface
}

type OrderBasedCrossover 

type OrderBasedCrossover struct {
	// contains filtered or unexported fields
}

Crossover for ordered chromosomes. Tends to preserve relative order.

parent1: A B C D E F parent2: d f b e a c

mask: * * _ * _ _ parent1: A B C D E F The genes 'A B D' will be copied in child in order from parent1 and in paces from parent2

child1 step1: A _ B _ D _

parent2: d f b e a c filler block: _ f _ e _ c

child1: A f B e D c Despite the fact that OrderBased behaves differently from Position crossover they are identical in expectation.

Source: Modeling Simple Genetic Algorithms for Permutation Problems. Darrell Whitley , Nam-wook Yoo (1995) http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.18.3585

func NewOrderBasedCrossover 

func NewOrderBasedCrossover() *OrderBasedCrossover

func (*OrderBasedCrossover) CanProduceCopiesOfParents 

func (crossover *OrderBasedCrossover) CanProduceCopiesOfParents(val bool) *OrderBasedCrossover

func (*OrderBasedCrossover) Crossover 

func (crossover *OrderBasedCrossover) Crossover(parents Chromosomes) Chromosomes

func (*OrderBasedCrossover) ParentsCount 

func (crossover *OrderBasedCrossover) ParentsCount() int

type OrderCrossoverVer1 

type OrderCrossoverVer1 struct{}

Crossover for ordered chromosomes. Tends to preserve relative order.

parent1: A B C D E parent2: d b e a c

cross section: _ * * _ _

child1 step1: _ B C _ _

parent2: d b e a c filler block: d _ e a _ The first element of filler block is added at the end of the cross setion.

child1: a B C d e

Source: Modeling Simple Genetic Algorithms for Permutation Problems. Darrell Whitley , Nam-wook Yoo (1995) http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.18.3585

type OrderCrossoverVer2 

type OrderCrossoverVer2 struct{}

Crossover for ordered chromosomes. Tends to preserve relative order.

parent1: A B C D E parent2: d b e a c

cross section: _ * * _ _

child1 step1: _ B C _ _

parent2: d b e a c filler block: d _ e a _ The first element of filler block is added at the start of the child.

child1: d B C e a

Source: On Genetic Crossover Operators for Relative Order Preservation http://www.dmi.unict.it/mpavone/nc-cs/materiale/moscato89.pdf

type OrderedChromosome 

type OrderedChromosome struct {
	*ChromosomeBase
	// contains filtered or unexported fields
}

func NewOrderedChromosome 

func NewOrderedChromosome(genes OrderedGenes) *OrderedChromosome

func (*OrderedChromosome) Genes 

func (chrom *OrderedChromosome) Genes() GenesInterface

func (*OrderedChromosome) Ind 

func (chrom *OrderedChromosome) Ind(val int) int

func (*OrderedChromosome) OrderedGenes 

func (chrom *OrderedChromosome) OrderedGenes() OrderedGenes

func (*OrderedChromosome) String 

func (chrom *OrderedChromosome) String() string

type OrderedGenes 

type OrderedGenes []int

func (OrderedGenes) Copy 

func (g OrderedGenes) Copy(genes GenesInterface, from1, from2, to2 int) int

func (OrderedGenes) Get 

func (g OrderedGenes) Get(i int) interface{}

func (OrderedGenes) Ind 

func (g OrderedGenes) Ind(val int) int

func (OrderedGenes) Len 

func (g OrderedGenes) Len() int

func (OrderedGenes) Set 

func (g OrderedGenes) Set(i int, val interface{})

func (OrderedGenes) Swap 

func (g OrderedGenes) Swap(i, j int)

type OrderedRandomInitializer 

type OrderedRandomInitializer struct {
}

func NewOrderedRandomInitializer 

func NewOrderedRandomInitializer() *OrderedRandomInitializer

func (*OrderedRandomInitializer) Init 

func (initializer *OrderedRandomInitializer) Init(count, chromSize int) Chromosomes

type PartiallyMappedCrossover 

type PartiallyMappedCrossover struct {
	// contains filtered or unexported fields
}

Crossover for ordered chromosomes. Tends to preserve absolute order.

Source: Modeling Simple Genetic Algorithms for Permutation Problems. Darrell Whitley , Nam-wook Yoo (1995) http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.18.3585

func NewPartiallyMappedCrossover 

func NewPartiallyMappedCrossover() *PartiallyMappedCrossover

func (*PartiallyMappedCrossover) CanProduceCopiesOfParents 

func (crossover *PartiallyMappedCrossover) CanProduceCopiesOfParents(val bool) *PartiallyMappedCrossover

func (*PartiallyMappedCrossover) Crossover 

func (crossover *PartiallyMappedCrossover) Crossover(parents Chromosomes) Chromosomes

func (*PartiallyMappedCrossover) ParentsCount 

func (crossover *PartiallyMappedCrossover) ParentsCount() int

type PositionBasedCrossover 

type PositionBasedCrossover struct {
	// contains filtered or unexported fields
}

Crossover for ordered chromosomes. Generalization of OrderCrossoverVer2. Tends to preserve relative order.

parent1: A B C D E parent2: d b e a c

mask: _ * _ * _ parent1: A B C D E

child1 step1: _ B _ D _

parent2: d b e a c filler block: _ _ e a c

child1: e B a D c

Source: Modeling Simple Genetic Algorithms for Permutation Problems. Darrell Whitley, Nam-wook Yoo (1995) http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.18.3585

func NewPositionBasedCrossover 

func NewPositionBasedCrossover() *PositionBasedCrossover

func (*PositionBasedCrossover) CanProduceCopiesOfParents 

func (crossover *PositionBasedCrossover) CanProduceCopiesOfParents(val bool) *PositionBasedCrossover

func (*PositionBasedCrossover) Crossover 

func (crossover *PositionBasedCrossover) Crossover(parents Chromosomes) Chromosomes

func (*PositionBasedCrossover) ParentsCount 

func (crossover *PositionBasedCrossover) ParentsCount() int

type PrecedencePreservativeCrossover 

type PrecedencePreservativeCrossover struct {
}

Crossover for ordered chromosomes. Tends to preserve absolute order.

Source: Introduction to Genetic Algorithms. S.N. Sivanandam, S. N. Deepa (2008) http://www.amazon.com/Introduction-Genetic-Algorithms-S-N-Sivanandam/dp/354073189X/

func NewPrecedencePreservativeCrossover 

func NewPrecedencePreservativeCrossover() *PrecedencePreservativeCrossover

func (*PrecedencePreservativeCrossover) Crossover 

func (crossover *PrecedencePreservativeCrossover) Crossover(parents Chromosomes) Chromosomes

func (*PrecedencePreservativeCrossover) ParentsCount 

func (crossover *PrecedencePreservativeCrossover) ParentsCount() int

type RelativeOrderingCrossover 

type RelativeOrderingCrossover struct {
	// contains filtered or unexported fields
}

Crossover for ordered chromosomes. Tends to preserve relative order.

Ordering genetic algorithms and deception. Hillol Kargupta, Ka. Lyanmoy Deb, David E. Goldberg (1992) http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.94.6805

func NewRelativeOrderingCrossover 

func NewRelativeOrderingCrossover(preservedGenes int) *RelativeOrderingCrossover

func (*RelativeOrderingCrossover) Crossover 

func (crossover *RelativeOrderingCrossover) Crossover(parents Chromosomes) Chromosomes

func (*RelativeOrderingCrossover) ParentsCount 

func (crossover *RelativeOrderingCrossover) ParentsCount() int

type RouletteWheelCostWeightingSelector 

type RouletteWheelCostWeightingSelector struct {
	*SelectorBase
	// contains filtered or unexported fields
}

Selects individual with probability proportional to it fitness value. Warning! In order to use this selector cost value must be normalized, i.e. chromosome with cost=0 is the best solution.

func NewRouletteWheelCostWeightingSelector 

func NewRouletteWheelCostWeightingSelector() *RouletteWheelCostWeightingSelector

func (*RouletteWheelCostWeightingSelector) Prepare 

func (selector *RouletteWheelCostWeightingSelector) Prepare(population Chromosomes)

func (*RouletteWheelCostWeightingSelector) SelectInd 

func (selector *RouletteWheelCostWeightingSelector) SelectInd() int

type RouletteWheelRankWeightingSelector 

type RouletteWheelRankWeightingSelector struct {
	*SelectorBase
	// contains filtered or unexported fields
}

Selects individual with probability proportional to it fitness value. Warning! In order to use this selector cost value must be normalized, i.e. chromosome with cost=0 is the best solution.

func NewRouletteWheelRankWeightingSelector 

func NewRouletteWheelRankWeightingSelector() *RouletteWheelRankWeightingSelector

func (*RouletteWheelRankWeightingSelector) Prepare 

func (selector *RouletteWheelRankWeightingSelector) Prepare(population Chromosomes)

func (*RouletteWheelRankWeightingSelector) SelectInd 

func (selector *RouletteWheelRankWeightingSelector) SelectInd() int

type SelectorBase 

type SelectorBase struct {
	SelectorBaseVirtualMInterface
	// contains filtered or unexported fields
}

Base class for selectors.

func NewSelectorBase 

func NewSelectorBase(virtual SelectorBaseVirtualMInterface) *SelectorBase

Constructor for SelectorBase

func (*SelectorBase) Prepare 

func (selector *SelectorBase) Prepare(population Chromosomes)

func (*SelectorBase) Select 

func (selector *SelectorBase) Select() ChromosomeInterface

func (*SelectorBase) SelectMany 

func (selector *SelectorBase) SelectMany(count int) Chromosomes

func (*SelectorBase) SelectManyAreUnique 

func (selector *SelectorBase) SelectManyAreUnique(value bool) *SelectorBase

Sets whether or not SelectMany will return only unique chromosomes

type SelectorBaseVirtualMInterface 

type SelectorBaseVirtualMInterface interface {
	SelectInd() int
}

SelectorBase's virtual methods

type SelectorInterface 

type SelectorInterface interface {
	// Called once before selection for each population
	Prepare(Chromosomes)

	// Selects one parent from population
	Select() ChromosomeInterface
	// Selects c parents from population
	SelectMany(c int) Chromosomes
}

Interface for selectors -

type SimpleOptimizer 

type SimpleOptimizer struct {
	*OptimizerBase
	// contains filtered or unexported fields
}

func NewSimpleOptimizer 

func NewSimpleOptimizer() *SimpleOptimizer

func (*SimpleOptimizer) CrossoverProbability 

func (optimizer *SimpleOptimizer) CrossoverProbability(crossoverProbability float64) *SimpleOptimizer

func (*SimpleOptimizer) Elitism 

func (optimizer *SimpleOptimizer) Elitism(elitism int) *SimpleOptimizer

type SimpleTournamentSelector 

type SimpleTournamentSelector struct {
	*SelectorBase
	// contains filtered or unexported fields
}

Selects several random individuals from population and then selectd best.

func NewRandomSelector 

func NewRandomSelector() *SimpleTournamentSelector

func NewSimpleTournamentSelector 

func NewSimpleTournamentSelector(contestants int) *SimpleTournamentSelector

Constructor for simple tournament selector. When contestants = 1 behaves like random selector

func (*SimpleTournamentSelector) SelectInd 

func (selector *SimpleTournamentSelector) SelectInd() int

type SimpleWeeder 

type SimpleWeeder struct {
	// contains filtered or unexported fields
}

SimpleWeeder weeds part of population proportional to rate.

func NewSimpleWeeder 

func NewSimpleWeeder(rate float64) *SimpleWeeder

Creates new SimpleWeeder. Rate must be in [0,100)

func (*SimpleWeeder) Weed 

func (weeder *SimpleWeeder) Weed(pop []ChromosomeInterface) []ChromosomeInterface

type StatisticsAggregatorConstructor 

type StatisticsAggregatorConstructor func(StatisticsOptionsInterface) StatisticsAggregatorInterface

type StatisticsAggregatorInterface 

type StatisticsAggregatorInterface interface {
	Aggregate(StatisticsDataInterface)
	// Calcs mean of aggregated data
	// Should be called on the end of aggregation
	Compute() StatisticsDataInterface
}

Statistics aggregator

func NewStatisticsDefaultAggregator 

func NewStatisticsDefaultAggregator(options StatisticsOptionsInterface) StatisticsAggregatorInterface

type StatisticsAggregatorWithOptions 

type StatisticsAggregatorWithOptions interface {
	StatisticsAggregatorInterface
	Options() StatisticsOptionsInterface
}

type StatisticsConstructor 

type StatisticsConstructor func(StatisticsOptionsInterface) StatisticsInterface

type StatisticsDataDefault 

type StatisticsDataDefault interface {
	Generations() int
	Duration() time.Duration
	Durations() *HierarchicalDuration
	MinCost() float64
	MinCosts() []float64
	GenerationsWithoutImprovements() int
	MinCostsVar() float64
	MeanCost() float64
	MeanCosts() []float64
	WorstCost() float64
	WorstCosts() []float64
}

type StatisticsDataInterface 

type StatisticsDataInterface interface{}

type StatisticsDefault 

type StatisticsDefault struct {
	// contains filtered or unexported fields
}

Default realization of StatisticsInterface

func (*StatisticsDefault) Data 

func (statistics *StatisticsDefault) Data() StatisticsDataInterface

func (*StatisticsDefault) Duration 

func (statistics *StatisticsDefault) Duration() time.Duration

Returns duration of optimization. If optimization are in progress then returns time elapsed from start.

func (*StatisticsDefault) Durations 

func (statistics *StatisticsDefault) Durations() *HierarchicalDuration

Returns entire tracked hierarchy of duration

func (*StatisticsDefault) End 

func (statistics *StatisticsDefault) End()

func (*StatisticsDefault) Generations 

func (statistics *StatisticsDefault) Generations() int

Number of generations

func (*StatisticsDefault) GenerationsWithoutImprovements 

func (statistics *StatisticsDefault) GenerationsWithoutImprovements() int

Number of generations during which the min cost remains unchanged

func (*StatisticsDefault) MeanCost 

func (statistics *StatisticsDefault) MeanCost() float64

Mean cost of last iteration

func (*StatisticsDefault) MeanCosts 

func (statistics *StatisticsDefault) MeanCosts() []float64

Mean cost of each iteration Len would be `Iterations() + 1` because of initial value

func (*StatisticsDefault) MinCost 

func (statistics *StatisticsDefault) MinCost() float64

Min cost for last iteration

func (*StatisticsDefault) MinCosts 

func (statistics *StatisticsDefault) MinCosts() []float64

Min cost for each iteration Len would be `Generations() + 1` because of initial value

func (*StatisticsDefault) MinCostsVar 

func (statistics *StatisticsDefault) MinCostsVar() float64

Variance of min costs Variance equals NaN until two different values of MinCost are obtained

func (*StatisticsDefault) OnGeneration 

func (statistics *StatisticsDefault) OnGeneration(population Chromosomes)

Expects sorted population

func (*StatisticsDefault) Start 

func (statistics *StatisticsDefault) Start(keys ...string)

func (*StatisticsDefault) WorstCost 

func (statistics *StatisticsDefault) WorstCost() float64

Worst cost for last iteration

func (*StatisticsDefault) WorstCosts 

func (statistics *StatisticsDefault) WorstCosts() []float64

Worst cost for each iteration Len would be `Iterations() + 1` because of initial value

type StatisticsDefaultAggregator 

type StatisticsDefaultAggregator struct {
	// contains filtered or unexported fields
}

func (*StatisticsDefaultAggregator) Aggregate 

func (aggregator *StatisticsDefaultAggregator) Aggregate(statistics StatisticsDataInterface)

func (*StatisticsDefaultAggregator) Compute 

func (aggregator *StatisticsDefaultAggregator) Compute() StatisticsDataInterface

func (*StatisticsDefaultAggregator) Data 

func (aggregator *StatisticsDefaultAggregator) Data() StatisticsDataInterface

func (*StatisticsDefaultAggregator) Duration 

func (aggregator *StatisticsDefaultAggregator) Duration() time.Duration

func (*StatisticsDefaultAggregator) Durations 

func (aggregator *StatisticsDefaultAggregator) Durations() *HierarchicalDuration

func (*StatisticsDefaultAggregator) Generations 

func (aggregator *StatisticsDefaultAggregator) Generations() int

func (*StatisticsDefaultAggregator) GenerationsWithoutImprovements 

func (aggregator *StatisticsDefaultAggregator) GenerationsWithoutImprovements() int

func (*StatisticsDefaultAggregator) MeanCost 

func (aggregator *StatisticsDefaultAggregator) MeanCost() float64

func (*StatisticsDefaultAggregator) MeanCosts 

func (aggregator *StatisticsDefaultAggregator) MeanCosts() []float64

func (*StatisticsDefaultAggregator) MinCost 

func (aggregator *StatisticsDefaultAggregator) MinCost() float64

func (*StatisticsDefaultAggregator) MinCosts 

func (aggregator *StatisticsDefaultAggregator) MinCosts() []float64

func (*StatisticsDefaultAggregator) MinCostsVar 

func (aggregator *StatisticsDefaultAggregator) MinCostsVar() float64

func (*StatisticsDefaultAggregator) Options 

func (aggregator *StatisticsDefaultAggregator) Options() StatisticsOptionsInterface

func (*StatisticsDefaultAggregator) WorstCost 

func (aggregator *StatisticsDefaultAggregator) WorstCost() float64

func (*StatisticsDefaultAggregator) WorstCosts 

func (aggregator *StatisticsDefaultAggregator) WorstCosts() []float64

type StatisticsDefaultOptions 

type StatisticsDefaultOptions struct {
	// contains filtered or unexported fields
}

Options for default statistics

func NewStatisticsDefaultOptions 

func NewStatisticsDefaultOptions() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) Copy 

func (options *StatisticsDefaultOptions) Copy() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) Ensure 

func (options *StatisticsDefaultOptions) Ensure(other StatisticsOptionsInterface)

func (*StatisticsDefaultOptions) TrackDurations 

func (options *StatisticsDefaultOptions) TrackDurations() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackGenerationsWithoutImprovements 

func (options *StatisticsDefaultOptions) TrackGenerationsWithoutImprovements() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackMeanCost 

func (options *StatisticsDefaultOptions) TrackMeanCost() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackMeanCosts 

func (options *StatisticsDefaultOptions) TrackMeanCosts() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackMinCosts 

func (options *StatisticsDefaultOptions) TrackMinCosts() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackMinCostsVar 

func (options *StatisticsDefaultOptions) TrackMinCostsVar() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackWorstCost 

func (options *StatisticsDefaultOptions) TrackWorstCost() *StatisticsDefaultOptions

func (*StatisticsDefaultOptions) TrackWorstCosts 

func (options *StatisticsDefaultOptions) TrackWorstCosts() *StatisticsDefaultOptions

type StatisticsInterface 

type StatisticsInterface interface {
	// Method for duration tracking
	// Call without parameters means that optimization is started
	Start(keys ...string)
	// Should be called when tracked operation is complete.
	End()
	// Optimizer will call this method on each generation
	// First call would be with initial population
	OnGeneration(Chromosomes)
	Data() StatisticsDataInterface
}

Accumulate statistics during optimization

func NewStatisticsDefault 

func NewStatisticsDefault(options StatisticsOptionsInterface) StatisticsInterface

type StatisticsOptionsInterface 

type StatisticsOptionsInterface interface {
	// Ensures that other tracks what ensurer tracks
	Ensure(other StatisticsOptionsInterface)
}

Options for statistics Defines what data gather and what not

type StopCriterionDefault 

type StopCriterionDefault struct {
	// contains filtered or unexported fields
}

func NewStopCriterionDefault 

func NewStopCriterionDefault() *StopCriterionDefault

func (*StopCriterionDefault) Max_Generations 

func (criterion *StopCriterionDefault) Max_Generations(value int) *StopCriterionDefault

Stop when specified number of generations is reached

func (*StopCriterionDefault) Max_GenerationsWithoutImprovements 

func (criterion *StopCriterionDefault) Max_GenerationsWithoutImprovements(value int) *StopCriterionDefault

Stop when min cost wasn't changed for value generations

func (*StopCriterionDefault) Min_Cost 

func (criterion *StopCriterionDefault) Min_Cost(value float64) *StopCriterionDefault

Stop when min cost less than or equals value

func (*StopCriterionDefault) Min_MinCostsVar 

func (criterion *StopCriterionDefault) Min_MinCostsVar(value float64) *StopCriterionDefault

Stop when variance of min costs less than or equals value

func (*StopCriterionDefault) Setup 

func (criterion *StopCriterionDefault) Setup(opts StatisticsOptionsInterface)

func (*StopCriterionDefault) ShouldStop 

func (criterion *StopCriterionDefault) ShouldStop(statistics StatisticsDataInterface) bool

type StopCriterionInterface 

type StopCriterionInterface interface {
	// Method sets up statistics tracking
	// Executes one time before optimization
	Setup(StatisticsOptionsInterface)

	// Method executes each time before iteration
	ShouldStop(StatisticsDataInterface) bool
}

type SwapMutator 

type SwapMutator struct {
}

Mutator simply swaps two elements

func (*SwapMutator) MutateCromosome 

func (mutator *SwapMutator) MutateCromosome(chrom ChromosomeInterface, ind int)

type TournamentSelector 

type TournamentSelector struct {
	*SelectorBase
	// contains filtered or unexported fields
}

Selects several random individuals from population and then runs tournament among them. See http://en.wikipedia.org/wiki/Tournament_selection

func NewTournamentSelector 

func NewTournamentSelector(probability float64, contestants int) *TournamentSelector

Constructor for tournament selector.

Probability specify the probability that the best individual will be selected. Must be equal or grater than 0.5

func (*TournamentSelector) SelectInd 

func (selector *TournamentSelector) SelectInd() int

type WeederInterface 

type WeederInterface interface {
	Weed([]ChromosomeInterface) []ChromosomeInterface
}

Weeder weeds part of population.

Source Files

View all Source files

Directories

Path Synopsis
examples
partition
partition_with_plot
Draws plot of min and mean costs versus generations.
 Draws plot of min and mean costs versus generations.
partition_with_plot_comparison
tsp
tsp_with_plot_comparison

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.