Python 遗传算法实现对原始数据的简单模拟

废话没有 直接上代码

import copy
import operator
import random
import math


class Genome:
    def __init__(self):
        self.fitness = 0
        self.gene_sequence = []


# 随机生成祖先群落
def gen_ancestors(value_array):
    ancestors = random.sample(value_array, 36)
    return ancestors


# 基因交换
def gene_change(array1, array2, pos, len):
    arr = array1[pos:len:1]
    del array1[pos:len]
    arr2 = array2[pos:len:1]
    del array2[pos:len]
    arr.reverse()
    arr2.reverse()

    for i in arr:
        array2.insert(pos, i)

    for i in arr2:
        array1.insert(pos, i)

    return array1, array2


# 基因突变
def gene_variation(array, pos):
    if array[pos] == 1:
        array[pos] = 0
    else:
        array[pos] = 1
    return array


# 基因适合度值判断
def survival_of_the_fittest(simple_array, new_gene_array):
    result_map = {}
    genomes = []
    for gene in new_gene_array:
        sum_of_squared_differences = 0
        index = 0

        for sige_gene in gene:
            sum_of_squared_differences += (sige_gene - simple_array[index]) * (sige_gene - simple_array[index])
            index += 1
        # print()
        genome = Genome()
        genome.fitness = format(math.sqrt(sum_of_squared_differences), '.4f')
        genome.gene_sequence = gene
        genomes.append(genome)

    def get_fitness(genome):
        return genome.fitness

    genomes.sort(key=get_fitness)

    # for genome in genomes:
    #     print(genome.fitness)
    # print(genome.gene_sequence)

    return copy.deepcopy(genomes[0:5])


# 数据展示
def reverse_flattening_display(raw_data, descendant_simulation):
    print("\n=========================原始数据展示========================")
    index = 0
    for i in raw_data:
        if index % 6 == 0:
            print()
        print(i, end='')
        index += 1
    pass
    print("\n=========================模拟数据展示========================")

    index = 0
    for i in descendant_simulation:
        if index % 6 == 0:
            print()
        print(i, end='')
        index += 1


if __name__ == '__main__':
    default_map = [[0, 1, 1, 1, 1, 0],
                   [0, 0, 0, 1, 0, 0],
                   [0, 1, 1, 1, 1, 0],
                   [0, 0, 0, 1, 0, 0],
                   [0, 1, 1, 1, 1, 0],
                   [0, 0, 0, 0, 0, 0]]

    value_array = []
    for values in default_map:
        for value in values:
            value_array.append(value)

    gen_ancestors_array = []
    for i in range(5):
        gen_ancestors_array.append(gen_ancestors(copy.deepcopy(value_array)))

    generation = 0
    while True:
        generation += 1
        first = random.randint(0, len(gen_ancestors_array) - 1)
        sec = -1
        while sec == first or sec == -1:
            sec = random.randint(0, len(gen_ancestors_array) - 1)
        arr1 = gen_ancestors_array[sec]
        arr2 = gen_ancestors_array[first]
        exanged_array = []
        exange_pos = random.randint(0, len(arr1) - 1)
        # print("基因交换位置： " + str(exange_pos))
        variation_pos = random.randint(0, len(arr1) - 1)
        # print("基因变异位置： " + str(variation_pos))

        arr3, arr4 = gene_change(copy.deepcopy(arr1), copy.deepcopy(arr2), exange_pos
                                 ,
                                 5)

        exanged_array.append(arr1)
        exanged_array.append(arr2)
        exanged_array.append(arr3)
        exanged_array.append(arr4)

        input_arr = gen_ancestors_array[random.randint(0, len(gen_ancestors_array) - 1)]
        arr_mutated = gene_variation(copy.deepcopy(input_arr),
                                     variation_pos)

        exanged_array.append(arr_mutated)

        # print("原图数据 = ", value_array)

        # print("==================================================")
        # for i in gen_ancestors_array:
        #     print("祖先数据 = ", i)

        # print("==================================================")

        # for i in exanged_array:
        #     print("子孙数据 = ", i)

        new_gene_array = []
        new_gene_array += copy.deepcopy(gen_ancestors_array)
        new_gene_array += copy.deepcopy(exanged_array)

        # for i in new_gene_array:
        #     print("后代全部基因 = ", i)

        superior_genes = survival_of_the_fittest(copy.deepcopy(value_array), copy.deepcopy(new_gene_array))
        reverse_flattening_display(value_array, superior_genes[0].gene_sequence)

        if operator.eq(value_array, superior_genes[0].gene_sequence):
            print("循环结束，共经历 " + str(generation) + " 代")
            break

        gen_ancestors_array.clear()
        superior_genes.reverse()
        for genes in superior_genes:
            gen_ancestors_array.append(genes.gene_sequence)
        # time.sleep(2)

运行结果 

=========================原始数据展示========================

011110
000100
011110
000100
011110
000000
=========================模拟数据展示========================

011110
000100
011110
000100
011110
000000循环结束，共经历 179 代

思路来自： 【数之道14】六分钟时间，带你走近遗传算法_哔哩哔哩_bilibili