def test_evolve_storage(self):
storage_instance = SqliteStorage("test.sqlite")
protogene = ProtoGene(IntGene, 'a', min_value=0, max_value=10)
protogenome = ProtoGenome([protogene], mutation_probability=0.1)
algo = GeneticAlgorithm(protogenome, fitness_evaluator,
termination_criteria=convergence_stop, storage_instance=storage_instance)
algo.evolve()
def test_3():
pop_size = ProtoGene(DiscreteGene, "pop_size", alleles=[100, 200, 300])
mutation_probability = ProtoGene(
DiscreteGene, "mutation_probability", alleles=[0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5]
)
num_parents = ProtoGene(DiscreteGene, "num_parents", alleles=[2, 4, 8, 16])
crossover_probability = ProtoGene(
DiscreteGene, "crossover_probability", alleles=[0.05, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]
)
crossover_method = ProtoGene(
DiscreteGene, "crossover_method", alleles=[single_point_crossover, genome_add, genome_choose]
)
protogenome = ProtoGenome([pop_size, mutation_probability, num_parents, crossover_probability, crossover_method])
algo = GeneticAlgorithm(
protogenome,
genetic_fitness,
population_size=10,
optimization_mode="min",
termination_criteria=convergence_stop,
termination_criteria_options={"num_generations": 5},
)
algo.evolve(debug=True)
bi = algo.best_individual()
print "*" * 10
print bi.dict_value()
return bi.score
def genetic_fitness(genome):
values = genome.dict_value()
pg_1 = ProtoGene(FloatGene, max_value=150.0, min_value=-150.0, mutation_max_step=5)
pg_2 = ProtoGene(FloatGene, max_value=150.0, min_value=-150.0, mutation_max_step=5)
protogenes = [pg_1, pg_2]
protogenome = ProtoGenome(protogenes, mutation_probability=values["mutation_probability"])
algo = GeneticAlgorithm(
protogenome,
schafferF6,
population_size=values["pop_size"],
optimization_mode="min",
num_parents=values["num_parents"],
# elitism = False,
crossover_method=values["crossover_method"],
crossover_probability=values["crossover_probability"],
selection_method=select_from_top,
termination_criteria=convergence_stop,
termination_criteria_options={"num_generations": 5},
)
algo.evolve()
bi = algo.best_individual()
print bi.score
return bi.score
def test_evaluate_population(self):
protogene = ProtoGene(IntGene, 'a', min_value=0, max_value=100)
protogenome = ProtoGenome([protogene])
algo = GeneticAlgorithm(protogenome, fitness_evaluator, termination_criteria=convergence_stop)
algo.initialize()
algo.evaluate_population()
for individual in algo.population.individuals:
assert individual.score == fitness_evaluator(individual)
def test_evolve(self):
protogene = ProtoGene(IntGene, 'a', min_value=0, max_value=10)
protogenome = ProtoGenome([protogene], mutation_probability=0.1)
algo = GeneticAlgorithm(
protogenome,
fitness_evaluator,
termination_criteria=[raw_score_stop,convergence_stop],
termination_criteria_options = [{'stop_score':0}]
)
algo.evolve()
def test_6():
#not yet converging!
return
protogenes = []
for square in board.data.keys():
row = square[0]
col = square[1]
value = board.data[(row, col)]
if value:
continue
name = str(row) + "_" + str(col)
alleles = range(1,10)
numbers_in_row = [x for x in board.get_row(row) if x ]
numbers_in_col = [x for x in board.get_col(col) if x ]
numbers_in_square = [x for x in board.get_square(row, col) if x]
alleles = [x for x in alleles if (x not in numbers_in_row)]
alleles = [x for x in alleles if (x not in numbers_in_col)]
alleles = [x for x in alleles if (x not in numbers_in_square)]
protogenes.append(ProtoGene(DiscreteGene, name, alleles=alleles))
protogenome = ProtoGenome(protogenes, mutation_probability=0.01)
algo = GeneticAlgorithm(protogenome, sudoku_fitness,
population_size = 200,
#num_populations=1,
#isolated_cycles=8,
optimization_mode = 'min',
num_parents = 2,
#elitism = False,
crossover_method = single_point_crossover,
crossover_probability=.5,
selection_method = simple_tournament_select,
termination_criteria=[raw_score_stop], termination_criteria_options=[{'stop_score':0}])
algo.evolve(debug=True)
bi = algo.best_individual()
print bi.dict_value()
def test_1():
protogenes = protogene_factory(IntGene, 'x', 50, min_value=0, max_value=10)
protogenome = ProtoGenome(protogenes, mutation_probability = 0.1)
algo = GeneticAlgorithm(protogenome, fitness_function,
population_size = 200,
optimization_mode = 'min',
num_parents = 4,
crossover_method = single_point_crossover,
selection_method = select_from_top,
termination_criteria=[raw_score_stop], termination_criteria_options=[{'stop_score':0}])
#main evolution cycle
algo.evolve(debug=True)
bi = algo.best_individual()
print bi.score
print bi.dict_value()
def test_5():
num_letters = len("genepi")
protogenes = protogene_factory(CharGene, 'x', num_letters)
protogenome = ProtoGenome(protogenes, mutation_probability = 0.05)
algo = GeneticAlgorithm(protogenome, findname,
population_size = 200,
optimization_mode = 'min',
num_parents = 8,
#elitism = False,
crossover_method = single_point_crossover,
crossover_probability = 0.5,
selection_method = select_from_top,
termination_criteria = raw_score_stop,
termination_criteria_options={'stop_score':0}
)
algo.evolve(debug=True)
bi = algo.best_individual()
print bi.score
print "".join(bi.list_value())
def test_evolve_population(self):
protogene = ProtoGene(IntGene, 'a', min_value=0, max_value=10)
protogenome = ProtoGenome([protogene])
algo = GeneticAlgorithm(protogenome, fitness_evaluator, termination_criteria=convergence_stop)
algo.initialize()
algo.evaluate_population()
g1 = algo.generation
algo.evolve_population()
g2 = algo.generation
assert g1 == 0 and g2 == 1
def test_2():
pg_1 = ProtoGene(FloatGene, max_value=150.0, min_value=-150.0, mutation_max_step=10)
pg_2 = ProtoGene(FloatGene, max_value=150.0, min_value=-150.0, mutation_max_step=10)
protogenes = [pg_1, pg_2]
protogenome = ProtoGenome(protogenes, mutation_probability = 0.15)
algo = GeneticAlgorithm(protogenome, schafferF6,
population_size = 200,
optimization_mode = 'min',
num_parents = 8,
#elitism = False,
crossover_method = single_point_crossover,
crossover_probability = 0.5,
selection_method = roulette_select,
termination_criteria = convergence_stop,
termination_criteria_options={'num_generations':20}
)
algo.evolve(debug=True)
bi = algo.best_individual()
print bi.score
print bi.dict_value()
def test_best_individual(self):
protogene = ProtoGene(IntGene, 'a', min_value=0, max_value=100)
protogenome = ProtoGenome([protogene])
algo = GeneticAlgorithm(protogenome, fitness_evaluator, termination_criteria=convergence_stop)
algo.initialize()
for i, individual in enumerate(algo.population.individuals):
individual.set_value('a', i)
algo.evaluate_population()
bi = algo.best_individual()
assert bi.score == 0
def test_should_terminate(self):
protogene = ProtoGene(IntGene, 'a', min_value=0, max_value=100, value=1)
protogenome = ProtoGenome([protogene], mutation_probability=1)
algo = GeneticAlgorithm(protogenome, fitness_evaluator, termination_criteria=convergence_stop)
algo.initialize()
for individual in algo.population.individuals:
individual.score = 1
for x in range(11):
stats = {'min_score' : 0}
algo.population_stats.append(stats)
algo.generation = 10
st = algo.should_terminate()
assert st == True
algo.generation = 1
st = algo.should_terminate()
assert st == False
def test_initialize(self):
algo = GeneticAlgorithm(self.protogenome, fitness_evaluator)
algo.initialize()
