def crossover(parent_one, parent_two):
if randint(0, 100) > crossover_rate * 100:
return (parent_one, parent_two)
midpoint = randint(1, task_count - 1)
offspring_one, offspring_two = Genome(False), Genome(False)
offspring_one.genes = parent_one.genes[:midpoint] + parent_two.genes[
midpoint:]
offspring_two.genes = parent_two.genes[:midpoint] + parent_one.genes[
midpoint:]
offspring_one.update_duration()
offspring_two.update_duration()
return (offspring_one, offspring_two)
def crossover(self, genome1, genome2):
#Once in a while genome2 is not set, pull request if you found why
if type(genome2) == type(None):
genome2 = randomGenome(self.genomes)
if Config.crossoverVersion == 1:
genome = Genome()
for i in range(genome1.size()):
# Crossover
if random.random() <= Config.uniformRate:
genome.setGene(i, genome1.getGene(i))
else:
genome.setGene(i, genome2.getGene(i))
return genome
#I know it's cryptic...
if Config.crossoverVersion == 2:
genome = Genome()
genes1 = genome1.genes
genes2 = genome2.genes
if random.random() < 0.5:
tmp = genes1
genes1 = genes2[:]
genes2 = tmp[:]
finalweights = genes1[:-genome1.network.numberOfBiases]
biases1 = genes1[genome1.network.numberOfWeights:]
biases2 = genes2[genome2.network.numberOfWeights:]
cutLocation = int(math.floor(len(biases1) * random.random()))
part1 = biases1[:cutLocation]
part2 = biases2[cutLocation:]
#Form the final genes
part1.extend(part2)
finalbiases = part1
finalweights.extend(finalbiases)
genome.genes = finalweights[:]
return genome