private static void PrintStatus(int generation, Chromosome[] population)
{
int minFitness = population.Min(ch => ch.Fitness);
int avgFitness = (int)population.Average(ch => ch.Fitness);
int maxFitness = population.Max(ch => ch.Fitness);
Console.WriteLine($"Generation {generation.PadLeft(3)}: Fitness; {minFitness.PadLeft(6)}; {avgFitness.PadLeft(6)}; {maxFitness.PadLeft(6)}");
}
private Chromosome[] CreateInitialPopulation()
{
var population = new Chromosome[args.PopulationSize];
for (int i = 0; i < args.PopulationSize; i++)
{
var gens = new BitArray(chromosomeLength);
for (int j = 0; j < chromosomeLength; j++)
{
if (random.NextDouble() > .5)
gens.Set(j, true);
}
population[i] = new Chromosome(gens);
EvaluateFitness(population[i]);
}
return population;
}
private Chromosome Cross(Chromosome chromosome, Chromosome other)
{
int randomCutIndex = random.Next(1, chromosomeLength);
var newGens = new BitArray(chromosome.Gens);
for (int i = randomCutIndex; i < chromosomeLength; i++)
{
newGens.Set(i, other.Gens[i]);
}
var offspring = new Chromosome(newGens);
if (random.NextDouble() < args.MutationProbability)
MutateRandomChromosome(offspring);
EvaluateFitness(offspring);
return offspring;
}
private void MutateRandomChromosome(Chromosome chromosome)
{
int randomGeneIndex = random.Next(0, chromosomeLength);
chromosome.Gens[randomGeneIndex] = !chromosome.Gens[randomGeneIndex];
}
private int FindWeakest(Chromosome[] population)
{
int minFitness = population[0].Fitness;
int weakestIndex = 0;
for (int j = 1; j < args.PopulationSize; j++)
{
if (population[j].Fitness < minFitness)
{
minFitness = population[j].Fitness;
weakestIndex = j;
}
}
return weakestIndex;
}
private List<Chromosome> SelectParents(Chromosome[] population, int parentsCount)
{
int populationFitness = population.Sum(ch => ch.Fitness);
breedingPool.Clear();
for (int i = 0; i < parentsCount; i++)
{
if (args.ParentSelection == ParentSelection.Tournament)
{
Chromosome fittest = null;
for (int j = 0; j < args.TournamentSize; j++)
{
int randomIndex = random.Next(0, args.PopulationSize);
var next = population[randomIndex];
if (fittest == null || fittest.Fitness < next.Fitness)
fittest = next;
}
breedingPool.Add(fittest);
}
else
{
int randomPoint = random.Next(0, populationFitness + 1);
int runningFitness = 0;
for (int j = 0; j < args.PopulationSize; j++)
{
runningFitness += population[j].Fitness;
if (runningFitness >= randomPoint)
{
breedingPool.Add(population[j]);
break;
}
}
}
}
return breedingPool;
}
private Chromosome[] CreateNewGeneration(Chromosome[] currentPopulation, out int elitesCount)
{
var newGeneration = new Chromosome[args.PopulationSize];
elitesCount = 0;
if (args.ElitesCount > 0)
{
var elites = currentPopulation.Where(ch => ch.Fitness > 0)
.OrderByDescending(ch => ch.Fitness)
.Take(args.ElitesCount).ToArray();
elitesCount = elites.Length;
Array.Copy(elites, newGeneration, elites.Length);
}
return newGeneration;
}
private bool GetVariableValue(int formulaIndex, Chromosome chromosome)
{
return currentFormula.Formula[formulaIndex].ToVariableValue(chromosome.Gens);
}
private void EvaluateFitness(Chromosome chromosome)
{
chromosome.Fitness = 0;
for (int i = 0; i < currentFormula.Formula.Length; i += 3)
{
var firstVar = GetVariableValue(i, chromosome);
var secondVar = GetVariableValue(i + 1, chromosome);
var thirdVar = GetVariableValue(i + 2, chromosome);
if (!(firstVar || secondVar || thirdVar))
return;
}
for (int i = 0; i < chromosomeLength; i++)
{
if (chromosome.Gens[i])
chromosome.Fitness += currentFormula.Weights[i];
}
}