给我一段最近插入的python代码解决TSP问题

以下是一个使用遗传算法解决TSP问题的Python代码示例：

import numpy as np
import random

# 创建数据集
num_cities = 10  # 城市数量
x = np.random.randint(0, 100, (num_cities, 2))  # 随机生成坐标
dist_matrix = np.zeros((num_cities, num_cities))  # 创建距离矩阵
for i in range(num_cities):
    for j in range(num_cities):
        dist_matrix[i, j] = np.sqrt(np.sum((x[i]-x[j])**2))  # 计算欧氏距离

# 定义遗传算法参数
pop_size = 50  # 种群大小
num_generations = 500  # 迭代次数
mutation_rate = 0.02  # 变异率

# 初始化种群
def init_population(pop_size):
    population = []
    for i in range(pop_size):
        chromosome = list(range(num_cities))
        random.shuffle(chromosome)
        population.append(chromosome)
    return population

# 计算适应度
def evaluate(chromosome):
    distance = 0
    for i in range(num_cities-1):
        distance += dist_matrix[chromosome[i], chromosome[i+1]]
    distance += dist_matrix[chromosome[-1], chromosome[0]]  # 回到起点
    return 1/distance

# 选择操作
def select(population):
    fitness = [evaluate(chromosome) for chromosome in population]
    fitness_sum = sum(fitness)
    select_prob = [f/fitness_sum for f in fitness]
    selected_idx = np.random.choice(range(pop_size), size=pop_size, replace=True, p=select_prob)
    return [population[idx] for idx in selected_idx]

# 交叉操作
def crossover(parent1, parent2):
    point1 = random.randint(0, num_cities-1)
    point2 = random.randint(0, num_cities-1)
    start = min(point1, point2)
    end = max(point1, point2)
    child = [-1] * num_cities
    for i in range(start, end+1):
        child[i] = parent1[i]
    idx = end
    for i in range(num_cities):
        if idx == num_cities:
            idx = 0
        if parent2[i] not in child:
            child[idx] = parent2[i]
            idx += 1
    return child

# 变异操作
def mutate(chromosome):
    if random.random() < mutation_rate:
        idx1 = random.randint(0, num_cities-1)
        idx2 = random.randint(0, num_cities-1)
        chromosome[idx1], chromosome[idx2] = chromosome[idx2], chromosome[idx1]

# 遗传算法主函数
def genetic_algorithm():
    population = init_population(pop_size)
    best_chromosome = []
    best_fitness = 0
    for _ in range(num_generations):
        population = select(population)
        offspring = []
        for i in range(int(pop_size/2)):
            parent1 = population[i]
            parent2 = population[pop_size-i-1]
            child1 = crossover(parent1, parent2)
            child2 = crossover(parent2, parent1)
            mutate(child1)
            mutate(child2)
            offspring.append(child1)
            offspring.append(child2)
        population = offspring
        fitness = [evaluate(chromosome) for chromosome in population]
        max_fitness = max(fitness)
        if max_fitness > best_fitness:
            best_fitness = max_fitness
            best_chromosome = population[fitness.index(max_fitness)]
    return best_chromosome, 1/best_fitness

# 运行遗传算法
best_chromosome, best_distance = genetic_algorithm()
print("最短路径: ", best_chromosome)
print("最短距离: ", best_distance)

这段代码使用遗传算法解决TSP问题，生成了一个包含10个城市的随机数据集，并输出了最短路径和最短距离。您可以根据需要修改数据集的大小和参数的值。