public void Start()
{
var fitness = new StringFitness(targetString);
var chromosome = new StringChromosome(targetString.Length);
var crossover = new UniformCrossover(0.5f);
var mutation = new UniformMutation();
var selection = new EliteSelection();
var population = new Population(1000, 1000, chromosome);
var termination = new FitnessThresholdTermination(1.00);
ga = new GeneticAlgorithm(population, fitness, selection, crossover, mutation);
ga.Termination = termination;
var latestFitness = 0.0;
ga.GenerationRan += (sender, e) =>
{
var bestChromosome = ga.BestChromosome as StringChromosome;
var bestFitness = bestChromosome.Fitness.Value;
Console.WriteLine(
"Generation {0}: \n{1} \nFitness: {2}",
ga.GenerationsNumber,
bestChromosome.Sentence,
bestFitness
);
if (bestFitness > latestFitness)
{
latestFitness = bestFitness;
}
};
ga.Start();
}
public ITermination FitnessThreshold()
{
var target = new FitnessThresholdTermination(1d);
var ga = Substitute.For <IGeneticAlgorithm>();
ga.BestChromosome.Returns(new TspChromosome(10)
{
Fitness = 1d
});
target.HasReached(ga);
return(target);
}
public void HasReached_BestChromosomeHasLowerFitness_False()
{
var repository = new MockRepository();
var ga = repository.StrictMock<IGeneticAlgorithm>();
using (repository.Ordered())
{
ga.Expect(g => g.BestChromosome).Return(new ChromosomeStub() { Fitness = 0.4 });
ga.Expect(g => g.BestChromosome).Return(new ChromosomeStub() { Fitness = 0.499 });
}
repository.ReplayAll();
var target = new FitnessThresholdTermination(0.5);
Assert.IsFalse(target.HasReached(ga));
Assert.IsFalse(target.HasReached(ga));
}
public void HasReached_BestChromosomeHasGreaterOrEqualFitness_True()
{
var ga = Substitute.For <IGeneticAlgorithm>();
ga.BestChromosome.Returns(new ChromosomeStub()
{
Fitness = 0.4
},
new ChromosomeStub()
{
Fitness = 0.8
});
var target = new FitnessThresholdTermination(0.8);
Assert.IsFalse(target.HasReached(ga));
Assert.IsTrue(target.HasReached(ga));
}
public static FloatingPointChromosome EvolveGeneticAlgorithm(FloatingPointChromosome chromosome, IOthelloAgent agent, string chromosomeLabel = "")
{
IPopulation population = new TplPopulation(30, 60, chromosome);
IFitness fitness = new EvaluationFitness(agent);
ISelection selection = new RouletteWheelSelection(); //Guess
ICrossover crossover = new UniformCrossover(); //Guess
IMutation mutation = new UniformMutation(); //Guess
ITermination stagnation = new FitnessStagnationTermination(500);
ITermination threshold = new FitnessThresholdTermination(.9);
ITaskExecutor taskExecutor = new ParallelTaskExecutor()
{
MaxThreads = Environment.ProcessorCount,
MinThreads = Environment.ProcessorCount / 2
};
GeneticAlgorithm algorithm = new GeneticAlgorithm(population, fitness, selection, crossover, mutation)
{
TaskExecutor = new TplTaskExecutor(),
MutationProbability = .2f
};
algorithm.TaskExecutor = taskExecutor;
algorithm.Termination = stagnation;
algorithm.Start();
SaveChromosome((FloatingPointChromosome)algorithm.BestChromosome, chromosomeLabel);
Debug.WriteLine("finished Training, with {0} time spent on evolving", algorithm.TimeEvolving);
Debug.WriteLine("fitness of this generation vs the last : {0}", algorithm.Fitness);
return((FloatingPointChromosome)algorithm.BestChromosome);
}
private static IAlgoritmo CriaAlgoritmoGenetico(Dictionary <string, string[]> dict, List <string> flat, Problema problema)
{
int populacaoMin, populacaoMax;
IPopulation population;
ISelection selection;
ICrossover crossover;
IMutation mutation;
ITermination termination;
IReinsertion reinsertion;
float crossoverProbability, mutationProbability;
var p = dict.ValueOrDefault("p", "50,100").Split(new[] { ',' });
if (p.Length != 2 || !int.TryParse(p[0], out populacaoMin) || !int.TryParse(p[1], out populacaoMax))
{
throw new ArgumentException("Faixa de população inválida.");
}
population = new Population(populacaoMin, populacaoMax, new CromossomoViajante(problema.Mapa.Locais.Count));
switch (dict.ValueOrDefault("s", "t"))
{
case "e":
selection = new EliteSelection();
break;
case "r":
selection = new RouletteWheelSelection();
break;
case "s":
selection = new StochasticUniversalSamplingSelection();
break;
case "t":
selection = new TournamentSelection();
break;
default:
throw new ArgumentException("Seleção inválida.");
}
switch (dict.ValueOrDefault("c", "o"))
{
case "s":
crossover = new CutAndSpliceCrossover();
break;
case "c":
crossover = new CycleCrossover();
break;
case "o":
crossover = new OrderedCrossover();
break;
case "ob":
crossover = new OrderBasedCrossover();
break;
case "op":
crossover = new OnePointCrossover();
break;
case "pm":
crossover = new PartiallyMappedCrossover();
break;
case "p":
crossover = new PositionBasedCrossover();
break;
case "tpa":
crossover = new ThreeParentCrossover();
break;
case "tp":
crossover = new TwoPointCrossover();
break;
case "u":
crossover = new UniformCrossover();
break;
default:
throw new ArgumentException("Crossover inválido.");
}
switch (dict.ValueOrDefault("m", "r"))
{
case "d":
mutation = new DisplacementMutation();
break;
case "f":
mutation = new FlipBitMutation();
break;
case "i":
mutation = new InsertionMutation();
break;
case "s":
mutation = new PartialShuffleMutation();
break;
case "r":
mutation = new ReverseSequenceMutation();
break;
case "t":
mutation = new TworsMutation();
break;
case "u":
mutation = new UniformMutation();
break;
default:
throw new ArgumentException("Mutação inválida.");
}
switch (dict.ValueOrDefault("t", "s"))
{
case "s":
termination = new FitnessStagnationTermination();
break;
case "t":
termination = new FitnessThresholdTermination();
break;
case "g":
termination = new GenerationNumberTermination();
break;
default:
throw new ArgumentException("Terminação inválida.");
}
switch (dict.ValueOrDefault("e", "e"))
{
case "e":
reinsertion = new ElitistReinsertion();
break;
case "p":
reinsertion = new PureReinsertion();
break;
case "u":
reinsertion = new UniformReinsertion();
break;
default:
throw new ArgumentException("Reinserção inválida.");
}
if (!float.TryParse(dict.ValueOrDefault("cp", "0,75"), out crossoverProbability))
{
throw new ArgumentException("Probabilidade de crossover inválida.");
}
if (!float.TryParse(dict.ValueOrDefault("mp", "0,25"), out mutationProbability))
{
throw new ArgumentException("Probabilidade de mutação inválida.");
}
return(new AlgoritmoGenetico(problema, population, selection, crossover, crossoverProbability, mutation, mutationProbability, termination, reinsertion));
}
public static void Main(string[] args)
{
//represents a sample in the population
var chromosome = new FloatingPointChromosome(
//min values array
Enumerable.Repeat(0.0, GoalString.Length).ToArray(),
//max values array
Enumerable.Repeat(25.0, GoalString.Length).ToArray(),
//bits nneded array
Enumerable.Repeat(5, GoalString.Length).ToArray(),
//decimals required array
Enumerable.Repeat(0, GoalString.Length).ToArray()
);
//create the population
var population = new Population(4, 8, chromosome);
//define a fitness function
var fitness = new FuncFitness((c) =>
{
var fc = c as FloatingPointChromosome;
var values = fc.ToFloatingPoints();
var numCorrect = 0;
for (int i = 0; i < GoalString.Length; i++)
{
var intVersion = Convert.ToInt32(values[i]);
if (intVersion == GoalString[i] - 'a')
{
numCorrect++;
}
}
return(Convert.ToInt64(Math.Pow(2, numCorrect)));
});
//select the top performers for reproduction
var selection = new EliteSelection();
//get half of genetic material from each parent
var crossover = new UniformCrossover(.5f);
//our numbers are internally represented as binary strings, randomly flip those bits
var mutation = new FlipBitMutation();
//stop mutating when there the goal is met (paried definition with fitness function)
var termination = new FitnessThresholdTermination(Math.Pow(2, GoalString.Length));
//put the genetic algorithm together
var ga = new GeneticAlgorithm(
population,
fitness,
selection,
crossover,
mutation);
ga.Termination = termination;
//print out the top performer of the population
ga.GenerationRan += (sender, e) =>
{
var bestChromosome = ga.BestChromosome as FloatingPointChromosome;
var bestFitness = Convert.ToInt32(bestChromosome.Fitness.Value);
if (bestFitness != HighestFitness)
{
HighestFitness = bestFitness;
}
var phenotype = bestChromosome.ToFloatingPoints();
var ints = phenotype.Select(Convert.ToInt32).ToArray();
var bestString = ConvertIntArrayToString(ints);
Console.WriteLine($"Best string: {bestString}");
};
ga.TerminationReached += (sender, eventArgs) => { Console.WriteLine("Finished Evolving"); };
ga.Start();
Console.ReadKey();
}
public GeneticOptimizer(GeneticOptimizerConfiguration configuration, Func <double[], double> objectiveFunction, Action <string> generationRanCallback = null)
{
//store configuration
_configuration = configuration;
_generationRanCallback = generationRanCallback;
//set min/max/precision of input variables
var minValues = new double[_configuration.Variables.Count];
var maxValues = new double[_configuration.Variables.Count];
var fractionDigits = new int[_configuration.Variables.Count];
for (int index = 0; index < _configuration.Variables.Count; index++)
{
minValues[index] = _configuration.Variables[index].MinimumValue;
maxValues[index] = _configuration.Variables[index].MaximumValue;
fractionDigits[index] = _configuration.Variables[index].NumberDigitsPrecision;
}
//total bits
var totalBits = new int[] { 64 };
//chromosome
var chromosome = new FloatingPointChromosome(minValues, maxValues, totalBits, fractionDigits);
//population
var population = new Population(MinimumNumberPopulation, MaximumNumberPopulation, chromosome);
//set fitness function
var fitnessFunction = new FuncFitness(c =>
{
var fc = c as FloatingPointChromosome;
var inputs = fc.ToFloatingPoints();
var result = objectiveFunction(inputs);
//add to results
if (!Double.IsNaN(result))
{
var list = inputs.ToList();
list.Add(result);
_result.IterationArray.Add(string.Join(",", list));
}
return(result);
});
//other variables
var selection = new EliteSelection();
var crossover = new UniformCrossover(0.5f);
var mutation = new FlipBitMutation();
var termination = new FitnessThresholdTermination();
_algorithm = new GeneticAlgorithm(population, fitnessFunction, selection, crossover, mutation)
{
Termination = termination,
};
//task parallelism
var taskExecutor = new ParallelTaskExecutor();
taskExecutor.MinThreads = 1;
taskExecutor.MaxThreads = _configuration.NumberThreadsToUse;
_algorithm.TaskExecutor = taskExecutor;
//if (_configuration.NumberThreadsToUse > 1)
//{
// var taskExecutor = new ParallelTaskExecutor();
// taskExecutor.MinThreads = 1;
// taskExecutor.MaxThreads = _configuration.NumberThreadsToUse;
// _algorithm.TaskExecutor = taskExecutor;
//}
//register generation ran callback
_algorithm.GenerationRan += AlgorithmOnGenerationRan;
}
