mep

package module
v0.0.0-...-9dd4931 Latest Latest
Warning

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

 Go to latest Published: Jul 1, 2016 License: MIT Imports: 9 Imported by: 0

README

go-mep

A pure Go implementation of the Multi Expression Programming algorithm

Alpha version, under heavy development.

TO DO:

  • Comprehensive tests
  • Main program
  • Finish test functions

Documentation

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

func ClassificationFF 

func ClassificationFF(signal, target []float64) float64

ClassificationFF -

func MeanErrorFF 

func MeanErrorFF(signal, target []float64) float64

MeanErrorFF -

func TotalErrorFF 

func TotalErrorFF(signal, target []float64) float64

TotalErrorFF -

Types

type CrossoverType 

type CrossoverType int

CrossoverType - uniform or onecutpoint 

const (
	// OneCutPoint - crossover type
	OneCutPoint CrossoverType = iota
	// Uniform - crossover type
	Uniform
)

type FitnessFunction 

type FitnessFunction func(signal, target []float64) float64

FitnessFunction -

type Mep 

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

Mep - primary class

func New 

func New(td TrainingData, ff FitnessFunction) *Mep

New - create a new Multi-Expression population

func (*Mep) Best 

func (m *Mep) Best() (float64, string)

Best - return the best fitness,expression of the population

func (*Mep) BestExpr 

func (m *Mep) BestExpr() string

BestExpr - return the best expression of the population

func (*Mep) BestFitness 

func (m *Mep) BestFitness() float64

BestFitness - return the best fitness of the population

func (*Mep) Evolve 

func (m *Mep) Evolve()

Evolve - one generation of population and sort for best fitness

func (*Mep) Oper 

func (m *Mep) Oper(all bool) []string

Oper - list of enabled operators

func (*Mep) PrintBest 

func (m *Mep) PrintBest()

PrintBest - print the best member of the population

func (*Mep) PrintTestData 

func (m *Mep) PrintTestData()

PrintTestData - print the testdata

func (*Mep) SetConst 

func (m *Mep) SetConst(fixed []float64, numRand int, minRand, maxRand float64)

SetConst - set fixed and random constants (resets population)

func (*Mep) SetCrossover 

func (m *Mep) SetCrossover(crossoverType CrossoverType, crossoverProbability float64)

SetCrossover - crossover type and probability (valid range 0.0 - 1.0)

func (*Mep) SetMutation 

func (m *Mep) SetMutation(mutationProbability float64)

SetMutation - mutation probability (valid range 0.0 - 1.0)

func (*Mep) SetOper 

func (m *Mep) SetOper(operName string, state bool)

SetOper - enable/disable operator

func (*Mep) SetPop 

func (m *Mep) SetPop(popSize, numSubpopulation, codeLength int)

SetPop - set population size and code length (resets population)

func (*Mep) SetProb 

func (m *Mep) SetProb(mutationProbability, crossoverProbability float64)

SetProb - set mutation/crossover probability (valid range 0.0 - 1.0)

func (*Mep) Solve 

func (m *Mep) Solve(numGens int, fitnessThreshold float64, showProgress bool) (int, time.Duration)

Solve - Evolve until fitnessThreshold or numGens is reached. Returns generations and total time

type TrainingData 

type TrainingData struct {
	Train  [][]float64
	Target []float64
	Labels []string
}

TrainingData -

func NewAckley 

func NewAckley(numTraining int) TrainingData

NewAckley -

func NewBooth 

func NewBooth(numTraining int) TrainingData

NewBooth -

func NewDejongF1 

func NewDejongF1(numTraining int) TrainingData

NewDejongF1 -

func NewDropwave 

func NewDropwave(numTraining int) TrainingData

NewDropwave -

func NewKepler 

func NewKepler(numTraining int) TrainingData

NewKepler -

func NewMichalewicz 

func NewMichalewicz(numTraining int) TrainingData

NewMichalewicz -

func NewPiTest 

func NewPiTest(numTraining int) TrainingData

NewPiTest -

func NewPythagorean 

func NewPythagorean(numTraining int) TrainingData

NewPythagorean -

func NewQuarticPoly 

func NewQuarticPoly(numTraining int) TrainingData

NewQuarticPoly -

func NewRastigrinF1 

func NewRastigrinF1(numTraining int) TrainingData

NewRastigrinF1 -

func NewRosenbrock 

func NewRosenbrock(numTraining int) TrainingData

NewRosenbrock -

func NewSchafferF6 

func NewSchafferF6(numTraining int) TrainingData

NewSchafferF6 -

func NewSchwefel 

func NewSchwefel(numTraining int) TrainingData

NewSchwefel -

func NewSequenceInduction 

func NewSequenceInduction(numTraining int) TrainingData

NewSequenceInduction -

func NewSimpleConstantRegression1 

func NewSimpleConstantRegression1(numTraining int) TrainingData

NewSimpleConstantRegression1 -

func NewSimpleConstantRegression2 

func NewSimpleConstantRegression2(numTraining int) TrainingData

NewSimpleConstantRegression2 -

func NewSimpleConstantRegression3 

func NewSimpleConstantRegression3(numTraining int) TrainingData

NewSimpleConstantRegression3 -

func NewSixHump 

func NewSixHump(numTraining int) TrainingData

NewSixHump -

func ReadTrainingData 

func ReadTrainingData(filename string, header bool, sep string) TrainingData

ReadTrainingData - read trainging data from a file

Source Files

View all Source files

Directories

Path Synopsis
Package is an implementation of the Multi Expression Programming algorithm Copyright 2016 Mark Chenoweth Licensed under terms of MIT license (see LICENSE) Usage: mep -h | -help mep -v | -version mep -o | -oper mep [options] <filename>|testdata Options: -h -help print help -v -version print version -o -oper print operators -td print testdata -summary print summary only -popsize=<subPopSize> sets sub-population size (default=100) -numpop=<numSubPop> sets number of sub-populations (default=1) -code=<codeLen> sets code length (default=50) -gens=<numGens> sets number of generations to evolve -seed=<int> sets random number seed (default=unixNano time) -fitness=<float> sets fitness threshold to stop evolving -mp=<mutationProb> sets mutation probability -cp=<crossoverProb> sets crossover probability -const=num,min,max sets random constant parameters (-const=num,min,max[,(e|pi|<fixed>)]) -enable=<op[,op]> enables operators (comma separated list) -disable=<op[,op]> disables operators (comma separated list)
 Package is an implementation of the Multi Expression Programming algorithm Copyright 2016 Mark Chenoweth Licensed under terms of MIT license (see LICENSE) Usage: mep -h | -help mep -v | -version mep -o | -oper mep [options] <filename>|testdata Options: -h -help print help -v -version print version -o -oper print operators -td print testdata -summary print summary only -popsize=<subPopSize> sets sub-population size (default=100) -numpop=<numSubPop> sets number of sub-populations (default=1) -code=<codeLen> sets code length (default=50) -gens=<numGens> sets number of generations to evolve -seed=<int> sets random number seed (default=unixNano time) -fitness=<float> sets fitness threshold to stop evolving -mp=<mutationProb> sets mutation probability -cp=<crossoverProb> sets crossover probability -const=num,min,max sets random constant parameters (-const=num,min,max[,(e|pi|<fixed>)]) -enable=<op[,op]> enables operators (comma separated list) -disable=<op[,op]> disables operators (comma separated list)

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.