遗传算法 生成位模式

xhz

生成位模式
以下示例显示了如何根据One Max问题生成一个包含15个字符串的位串。
如下所示导入必要的软件包 -

import random
from deap import base, creator, tools

定义评估函数。 这是创建遗传算法的第一步。

def eval_func(individual):
   target_sum = 15
   return len(individual) - abs(sum(individual) - target_sum)

现在，使用正确的参数创建工具箱 -

def create_toolbox(num_bits):
   creator.create("FitnessMax", base.Fitness, weights=(1.0,))
   creator.create("Individual", list, fitness=creator.FitnessMax)

初始化工具箱

toolbox = base.Toolbox()
toolbox.register("attr_bool", random.randint, 0, 1)
toolbox.register("individual", tools.initRepeat, creator.Individual,
   toolbox.attr_bool, num_bits)
toolbox.register("population", tools.initRepeat, list, toolbox.individual)

注册计算操作符 -

toolbox.register("evaluate", eval_func)

现在，注册交叉运算符 -

toolbox.register("mate", tools.cxTwoPoint)

注册一个可变运算符 -

toolbox.register("mutate", tools.mutFlipBit, indpb = 0.05)

定义育种操作符 -

toolbox.register("select", tools.selTournament, tournsize = 3)
return toolbox
if __name__ == "__main__":
   num_bits = 45
   toolbox = create_toolbox(num_bits)
   random.seed(7)
   population = toolbox.population(n = 500)
   probab_crossing, probab_mutating = 0.5, 0.2
   num_generations = 10
   print('\nEvolution process starts')

评估整个人口 -

fitnesses = list(map(toolbox.evaluate, population))
for ind, fit in zip(population, fitnesses):
   ind.fitness.values = fit
print('\nEvaluated', len(population), 'individuals')

经过几代人的创建和迭代 -

for g in range(num_generations):
   print("\n- Generation", g)

选择下一代个人 -

offspring = toolbox.select(population, len(population))

现在，克隆选定的个人 -

offspring = list(map(toolbox.clone, offspring))

对后代应用交叉和变异 -

for child1, child2 in zip(offspring[::2], offspring[1::2]):
   if random.random() < probab_crossing:
   toolbox.mate(child1, child2)

删除孩子的适应值

del child1.fitness.values
del child2.fitness.values

现在，应用突变 -

for mutant in offspring:
   if random.random() < probab_mutating:
   toolbox.mutate(mutant)
   del mutant.fitness.values

评估与无效的健身个体 -

invalid_ind = [ind for ind in offspring if not ind.fitness.valid]
fitnesses = map(toolbox.evaluate, invalid_ind)
for ind, fit in zip(invalid_ind, fitnesses):
   ind.fitness.values = fit
print('Evaluated', len(invalid_ind), 'individuals')

现在，用下一代个体替代人口 -

population[:] = offspring

打印当代人的统计数据 -

fits = [ind.fitness.values[0] for ind in population]
length = len(population)
mean = sum(fits) / length
sum2 = sum(x*x for x in fits)
std = abs(sum2 / length - mean**2)**0.5
print('Min =', min(fits), ', Max =', max(fits))
print('Average =', round(mean, 2), ', Standard deviation =',
round(std, 2))
print("\n- Evolution ends")

打印最终输出 -

   best_ind = tools.selBest(population, 1)[0]
   print('\nBest individual:\n', best_ind)
   print('\nNumber of ones:', sum(best_ind))
Following would be the output:
Evolution process starts
Evaluated 500 individuals
- Generation 0
Evaluated 295 individuals
Min = 32.0 , Max = 45.0
Average = 40.29 , Standard deviation = 2.61
- Generation 1
Evaluated 292 individuals
Min = 34.0 , Max = 45.0
Average = 42.35 , Standard deviation = 1.91
- Generation 2
Evaluated 277 individuals
Min = 37.0 , Max = 45.0
Average = 43.39 , Standard deviation = 1.46
… … … …
- Generation 9
Evaluated 299 individuals
Min = 40.0 , Max = 45.0
Average = 44.12 , Standard deviation = 1.11
- Evolution ends
Best individual:
[0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 1,
1, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0,
1, 0, 0, 1, 1, 1, 0, 0, 1, 0, 1]
Number of ones: 15


//更多请阅读：https://www.yiibai.com/ai_with_python/ai_with_python_genetic_algorithms.html