def test_should_the_solution_remain_unchanged_if_the_probability_is_zero(self):
operator = Polynomial(0.0)
solution = FloatSolution(2, 1, 0, [-5, -5, -5], [5, 5, 5])
solution.variables = [1.0, 2.0, 3.0]
mutated_solution = operator.execute(solution)
self.assertEqual([1.0, 2.0, 3.0], mutated_solution.variables)
def test_should_the_solution_change__if_the_probability_is_one(self):
operator = Polynomial(1.0)
solution = FloatSolution(2, 1, 0, [-5, -5, -5], [5, 5, 5])
solution.variables = [1.0, 2.0, 3.0]
FloatSolution.lower_bound = [-5, -5, -5]
FloatSolution.upper_bound = [5, 5, 5]
mutated_solution = operator.execute(solution)
self.assertNotEqual([1.0, 2.0, 3.0], mutated_solution.variables)
def main() -> None:
class NSGA2b(NSGAII[S, R]):
def is_stopping_condition_reached(self):
# Re-define the stopping condition
reached = [False, True][self.get_current_computing_time() > 4]
if reached:
logger.info("Stopping condition reached!")
return reached
problem = Fonseca()
algorithm = NSGA2b[FloatSolution, List[FloatSolution]](
problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0 / problem.number_of_variables,
distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
selection=BinaryTournament(RankingAndCrowdingDistanceComparator()))
algorithm.run()
result = algorithm.get_result()
SolutionListOutput[FloatSolution].print_function_values_to_file(
"FUN." + problem.get_name(), result)
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
def main() -> None:
problem = ZDT1()
algorithm = NSGAII[FloatSolution, List[FloatSolution]](
problem=problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0 / problem.number_of_variables,
distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
#selection=BinaryTournamentSelection(RankingAndCrowdingDistanceComparator()))
selection=BinaryTournament2Selection([
SolutionAttributeComparator("dominance_ranking"),
SolutionAttributeComparator("crowding_distance",
lowest_is_best=False)
]))
algorithm.run()
result = algorithm.get_result()
SolutionListOutput[FloatSolution].print_function_values_to_file(
"FUN." + problem.get_name(), result)
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
logger.info("Computing time: " + str(algorithm.total_computing_time))
def main() -> None:
class GGA2(GenerationalGeneticAlgorithm[FloatSolution, FloatSolution]):
def is_stopping_condition_reached(self):
# Re-define the stopping condition
reached = [False, True][self.get_current_computing_time() > 4]
if reached:
logger.info("Stopping condition reached!")
return reached
variables = 10
problem = Sphere(variables)
algorithm = GGA2(problem,
population_size=100,
max_evaluations=0,
mutation=Polynomial(1.0 / variables,
distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
selection=BinaryTournamentSelection())
algorithm.run()
result = algorithm.get_result()
logger.info("Algorithm (stop for timeout): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
logger.info("Solution: " + str(result.variables))
logger.info("Fitness: " + str(result.objectives[0]))
def main() -> None:
variables = 10
problem = Sphere(variables)
algorithm = NonElitistEvolutionStrategy[FloatSolution, FloatSolution]\
(problem, mu=10, lambdA=10, max_evaluations= 50000, mutation=Polynomial(1.0/variables))
algorithm.run()
result = algorithm.get_result()
print("Algorithm: " + algorithm.get_name())
print("Problem: " + problem.get_name())
print("Solution: " + str(result.variables))
print("Fitness: " + str(result.objectives[0]))
print("Computing time: " + str(algorithm.total_computing_time))
def main() -> None:
problem = Kursawe()
algorithm = NSGAII[FloatSolution, List[FloatSolution]](
problem=problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0/problem.number_of_variables, distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
selection = BinaryTournament(RankingAndCrowdingDistanceComparator()))
algorithm.run()
result = algorithm.get_result()
SolutionListOutput[FloatSolution].print_function_values_to_file("FUN."+problem.get_name(), result)
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
def main() -> None:
problem = Kursawe()
algorithm = SMPSO(problem=problem,
swarm_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0 / problem.number_of_variables,
distribution_index=20),
leaders=CrowdingDistanceArchive(100))
algorithm.run()
result = algorithm.get_result()
SolutionListOutput[FloatSolution].print_function_values_to_file(
"FUN." + problem.get_name(), result)
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
def main() -> None:
variables = 10
problem = Sphere(variables)
algorithm = GenerationalGeneticAlgorithm[FloatSolution, FloatSolution](
problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0/variables, distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
selection=BinaryTournamentSelection())
algorithm.run()
result = algorithm.get_result()
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
logger.info("Solution: " + str(result.variables))
logger.info("Fitness: " + str(result.objectives[0]))
def main() -> None:
problem = ZDT1()
algorithm = NSGAII[FloatSolution, List[FloatSolution]](
problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0/problem.number_of_variables, distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
#selection=BinaryTournament(RankingAndCrowdingDistanceComparator()))
selection = BinaryTournament2([SolutionAttributeComparator("dominance_ranking"),
SolutionAttributeComparator("crowding_distance", lowest_is_best=False)]))
observer = AlgorithmObserver(animation_speed=1*10e-8)
algorithm.observable.register(observer=observer)
algorithm.run()
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
def main() -> None:
variables = 10
problem = Sphere(variables)
algorithm = ElitistEvolutionStrategy[FloatSolution, FloatSolution](
problem=problem,
mu=10,
lambd_a=10,
max_evaluations=50000,
mutation=Polynomial(probability=1.0 / variables))
algorithm.start()
print("Algorithm (running as a thread): " + algorithm.get_name())
print("Problem: " + problem.get_name())
algorithm.join()
result = algorithm.get_result()
print("Solution: " + str(result.variables))
print("Fitness: " + str(result.objectives[0]))
print("Computing time: " + str(algorithm.total_computing_time))
def main() -> None:
variables = 10
problem = Sphere(variables)
algorithm = GenerationalGeneticAlgorithm[FloatSolution, FloatSolution](
problem=problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(probability=1.0 / variables,
distribution_index=20),
crossover=SBX(probability=1.0, distribution_index=20),
selection=BinaryTournamentSelection())
algorithm.run()
result = algorithm.get_result()
print("Algorithm: " + algorithm.get_name())
print("Problem: " + problem.get_name())
print("Solution: " + str(result.variables))
print("Fitness: " + str(result.objectives[0]))
print("Computing time: " + str(algorithm.total_computing_time))
def main() -> None:
variables = 10
problem = Sphere(variables)
algorithm = GenerationalGeneticAlgorithm[FloatSolution, FloatSolution](
problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0 / variables, distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
selection=BinaryTournament())
observer = BasicAlgorithmConsumer(2000)
algorithm.observable.register(observer=observer)
algorithm.start()
algorithm.join()
result = algorithm.get_result()
logger.info("Algorithm: " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
logger.info("Solution: " + str(result.variables))
logger.info("Fitness: " + str(result.objectives[0]))
def main() -> None:
problem = ZDT1()
algorithm = NSGAII[FloatSolution, List[FloatSolution]](
problem,
population_size = 100,
max_evaluations = 25000,
mutation = Polynomial(1.0/problem.number_of_variables, distribution_index=20),
crossover = SBX(1.0, distribution_index=20),
selection = BinaryTournament(RankingAndCrowdingDistanceComparator()))
observer = BasicAlgorithmConsumer(1000)
algorithm.observable.register(observer=observer)
algorithm.run()
result = algorithm.get_result()
SolutionListOutput[FloatSolution].print_function_values_to_file("FUN."+problem.get_name(), result)
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
logger.info("Computing time: " + str(algorithm.total_computing_time))
def main() -> None:
problem = Kursawe()
algorithm = NSGAII[FloatSolution, List[FloatSolution]](
problem,
population_size=100,
max_evaluations=25000,
mutation=Polynomial(1.0 / problem.number_of_variables,
distribution_index=20),
crossover=SBX(1.0, distribution_index=20),
selection=BinaryTournamentSelection())
algorithm.run()
result = algorithm.get_result()
SolutionListOutput[FloatSolution].plot_scatter_to_file(result,
file_name="FUN." +
problem.get_name(),
output_format='eps',
dpi=200)
SolutionListOutput[FloatSolution].plot_scatter_to_screen(result)
logger.info("Algorithm (continuous problem): " + algorithm.get_name())
logger.info("Problem: " + problem.get_name())
def test_should_constructor_create_a_valid_object(self):
problem = self.__DummyFloatProblem()
algorithm = SMPSO(problem=problem,
swarm_size=100,
max_evaluations=200,
mutation=Polynomial(probability=1.0 /
problem.number_of_variables),
leaders=BoundedArchive[FloatSolution](100))
self.assertEqual(1.5, algorithm.c1_min)
self.assertEqual(2.5, algorithm.c1_max)
self.assertEqual(1.5, algorithm.c2_min)
self.assertEqual(2.5, algorithm.c2_max)
self.assertEqual(0.1, algorithm.min_weight)
self.assertEqual(0.1, algorithm.max_weight)
self.assertEqual(-1.0, algorithm.change_velocity1)
self.assertEqual(-1.0, algorithm.change_velocity2)
self.assertEqual(200, algorithm.max_evaluations)
self.assertEqual((100, 2), algorithm.speed.shape)
numpy.testing.assert_array_almost_equal(numpy.array([2.0, 2.0]),
algorithm.delta_max)
numpy.testing.assert_array_almost_equal(algorithm.delta_max * -1.0,
algorithm.delta_min)
def test_should_constructor_create_a_non_null_object(self):
mutation = Polynomial(1.0)
self.assertIsNotNone(mutation)
def test_should_constructor_create_a_valid_operator(self):
operator = Polynomial(0.5, 20)
self.assertEqual(0.5, operator.probability)
self.assertEqual(20, operator.distribution_index)
def test_should_constructor_raise_an_exception_if_the_probability_is_greater_than_one(self):
with self.assertRaises(Exception):
Polynomial(2)
def test_should_constructor_raise_an_exception_if_the_probability_is_lower_than_zero(self):
with self.assertRaises(Exception):
Polynomial(-12)
