public InnovationPool(Monsters monsters, Map currentMap, Delegate callback)
{
this.monsters = monsters;
this.callback = callback;
originalMap = currentMap.CloneJson() as Map;
InitialPopulation = 30;
GenerationLimit = 30;
MutationPercentage = 0.35;
CrossOverPercentage = 0.4;
ElitismPercentage = 10;
running = false;
HasSolution = false;
cells = originalMap.SpawnCells;
Chromosome chrom = ChromosomeUtils.ChromosomeFromMap(originalMap);
string binaryString = chrom.ToBinaryString();
fitness = new ConvergenceFitness(cells, binaryString);
//we can create an empty population as we will be creating the
//initial solutions manually.
population = new Population(InitialPopulation, cells.Count * ChromosomeUtils.NUMBER_GENES, true, true, ParentSelectionMethod.FitnessProportionateSelection);
}
public ConvergencePool(Monsters monsters, Map currentMap, Delegate callback)
{
this.monsters = monsters;
this.callback = callback;
originalMap = currentMap.CloneJson() as Map;
InitialPopulation = 30;
GenerationLimit = 30;
MutationPercentage = 0.4;
CrossOverPercentage = 0.8;
ElitismPercentage = 10;
running = false;
HasSolution = false;
cells = originalMap.SpawnCells;
Chromosome chrom = ChromosomeUtils.ChromosomeFromMap(originalMap);
string binaryString = chrom.ToBinaryString();
fitness = new ConvergenceFitness(cells, binaryString);
population = new Population(InitialPopulation, cells.Count * ChromosomeUtils.NUMBER_GENES, true, true, ParentSelectionMethod.FitnessProportionateSelection);
population.Solutions.Clear();
for (int i = 0; i < 5; i++)
{
population.Solutions.Add(new Chromosome(binaryString));
}
for (int i = 0; i < 25; i++)
{
population.Solutions.Add(new Chromosome(cells.Count * ChromosomeUtils.NUMBER_GENES));
}
}
internal void Run(AreaManager areaManager, Population conv, Population inno, Population obj)
{
if (running)
{
return;
}
cells = originalMap.SpawnCells;
Chromosome chrom = ChromosomeUtils.ChromosomeFromMap(originalMap);
string binaryString = chrom.ToBinaryString();
convergenceFitness = new ConvergenceFitness(cells, binaryString);
guidelineFitness = new Guideline(cells, areaManager, MaxMonsters, MaxItens, HordesPercentage)
{
MaxItemsLever = LeverMaxItem,
MaxMonstersLever = LeverMaxMonster,
AmountHordesLever = LeverAmountHordes,
DangerLever = LeverDanger,
AccessibilityLever = LeverAccessibility
};
double total = GuidelinePercentage + UserPercentage + InnovationPercentage;
guidelineLevel = GuidelinePercentage / total;
ConvergenceLevel = UserPercentage / total;
innovationLevel = InnovationPercentage / total;
//we can create an empty population as we will be creating the
//initial solutions manually.
var population = new Population(InitialPopulation, cells.Count * ChromosomeUtils.NUMBER_GENES, true, true, ParentSelectionMethod.StochasticUniversalSampling);
population.Solutions.Clear();
population.Solutions.AddRange(obj.GetTop(10));
population.Solutions.AddRange(inno.GetTop(10));
population.Solutions.AddRange(conv.GetTop(10));
//create the elite operator
var elite = new Elite(ElitismPercentage);
//create the crossover operator
var crossover = new CrossoverIndex(CrossOverPercentage, ChromosomeUtils.NUMBER_GENES, true, GAF.Operators.CrossoverType.DoublePoint, ReplacementMethod.GenerationalReplacement);
//create the mutation operator
var mutate = new BinaryMutate(MutationPercentage);
//create the GA
var ga = new GeneticAlgorithm(population, CalculateFitness);
//add the operators
ga.Operators.Add(elite);
ga.Operators.Add(crossover);
//ga.Operators.Add(mutate);
ga.OnRunComplete += OnRunComplete;
//run the GA
running = true;
ga.Run(TerminateFunction);
}
