def _mutation(self):
self.mating_individuals = [
self.Individual(
mutate(x, self.dims, self.mutation_rate, self.mutation_eta))
for x in self.mating_individuals
]
for ind in self.mating_individuals:
ind.known_objectives = False
def generate(self, pop):
selected_parents = [x.value for x in random.sample(pop, 2)]
child = crossover(selected_parents[0], selected_parents[1], self.dims,
self.crossover_rate, self.crossover_eta)
return Individual(
self.trim_function(
mutate(child, self.dims, self.mutation_rate,
self.mutation_eta)))
def generate(self, pop):
selected_parents = [x.value for x in random.sample(pop, 2)]
child = crossover(selected_parents[0],
selected_parents[1],
self.dims,
self.crossover_rate,
self.crossover_eta)
return Individual(self.trim_function(mutate(child,
self.dims,
self.mutation_rate,
self.mutation_eta)))
def population_generator(self):
cost = 0
while True:
cost = self.calculate_fitnesses(self.__population, self.__archive)
self.__archive = self.environmental_selection(
self.__population, self.__archive)
self.population = [
self.trim_function(
mutate(
crossover(self.select(self.__archive),
self.select(self.__archive), self.dims,
self.crossover_rate, self.crossover_eta),
self.dims, self.mutation_rate, self.mutation_eta))
for _ in self.__population
]
yield SPEA2ImgaProxy(self, self.__archive, self.__population, cost)
def step(self):
self.cost += self.calculate_fitnesses(self.individuals, self.archive)
self.archive = self.environmental_selection(self.individuals,
self.archive)
self.population = [
self.trim_function(
mutate(
crossover(
self.select(self.archive),
self.select(self.archive),
self.dims,
self.crossover_rate,
self.crossover_eta,
),
self.dims,
self.mutation_rate,
self.mutation_eta,
)) for _ in self.individuals
]
def population_from_delegate(delegate, size, dims, rate, eta):
population = [[x for x in delegate]]
for _ in range(size - 1):
population.append(drivertools.mutate(delegate, dims, rate, eta))
return population
def _mutation(self):
self.mating_individuals = [
self.Individual(mutate(x, self.dims, self.mutation_rate, self.mutation_eta)) for x in
self.mating_individuals]
for ind in self.mating_individuals:
ind.known_objectives = False
def _mutation(self):
self.mating_individuals = [
Individual(
mutate(x, self.dims, self.mutation_rate, self.mutation_eta))
for x in self.mating_individuals
]
def _mutation(self):
self.mating_individuals = [
self.Individual(mutate(x, self.dims, self.mutation_rate, self.mutation_eta)) for x in
self.mating_individuals]
def mutate(self, xs):
return mutate(xs, self.dims, self.mutation_probability, self.mutation_variance)
def population_from_delegate_mutate(delegate, size, dims, rate, eta):
population = [[x for x in delegate]]
for _ in range(size - 1):
population.append(drivertools.mutate(delegate, dims, rate, eta))
return population