用python写一个解决tsp问题的算法

好的，下面是一个简单的TSP问题的近似算法，使用Python语言实现。

import numpy as np
import random

# TSP问题的近似算法
def tsp_approx(dists, start_city=None):
    """
    :param dists: 二维数组，表示城市之间的距离矩阵
    :param start_city: 起始城市编号，默认为0
    :return: 最短路径长度，最短路径的顺序
    """
    num_cities = len(dists)
    if start_city is None:
        start_city = random.randint(0, num_cities-1)
    cur_city = start_city
    unvisited_cities = set(range(num_cities))
    unvisited_cities.remove(cur_city)
    path = [cur_city]
    total_dist = 0

    while unvisited_cities:
        next_city = min(unvisited_cities, key=lambda x: dists[cur_city][x])
        total_dist += dists[cur_city][next_city]
        cur_city = next_city
        unvisited_cities.remove(cur_city)
        path.append(cur_city)
    total_dist += dists[path[-1]][start_city]
    return total_dist, path

这个算法使用贪心策略，每次选择距离当前城市最近的未访问城市。这种方法不能保证得到最优解，但是可以得到一个较好的近似解。