python编写的扫雷游戏

使用python语言编写的命令行方式的交互的《扫雷》游戏：

import numpy as np
import random
from typing import List

H = 36
W = 64

def belong_to(h, w, H=H, W=W):
    near = []
    for i in range(h-2, h+3):
        for j in range(w-2, w+3):
            if i>=0 and j>=0 and i<H and j<W and (i,j)!=(h,w):
                near.append((i, j))
    return near

def near_by(h, w, H=H, W=W):
    near = []
    for i in range(h-1, h+2):
        for j in range(w-1, w+2):
            if i>=0 and j>=0 and i<H and j<W and (i,j)!=(h,w):
                near.append((i,j))
    return near

def mine_count(h, w, real_state:np.array, H=H, W=W):
    count = 0
    for i, j in near_by(h, w, H=H, W=W):
        if real_state[i][j]==1:
            count += 1
    return count


class Env():
    def __init__(self, H=H, W=W, N=100):
        self.H = H
        self.W = W
        self.N = N

        # real state中0表示无雷，1表示有雷
        self.real_state = np.zeros((H, W), dtype=np.int32)
        self.mine = set()
        while len(self.mine)!=N:
            self.mine.add(random.randint(0, H*W))
        for x in self.mine:
            self.real_state[x//self.W][x%self.W] = 1    
            
        # state_type中0表示无雷，1-8表示有雷, 用此来表示对附近雷的计数
        self.state_type = np.zeros((H, W), dtype=np.int32)  
        for i in range(H):
            for j in range(W):
                self.state_type[i][j] = mine_count(h=i, w=j, H=H, W=W, real_state=self.real_state)

        # obs为-100表示未翻开(未知)，0-8表示翻开但无雷，数值大小表示翻开位置周边雷的数量
        # agent的状态记录所用，也可以用来作为打印之用
        self.obs = np.zeros((H, W), dtype=np.int32) -100
    
    def act(self, i, j):
        done = False
        if self.obs[i][j]!=-100:
            return ValueError
        if self.real_state[i][j] == 1:
            # game over 触雷
            done = True
            return None, done

        self.obs[i][j] = self.state_type[i][j]
        return self.obs[i][j], done
        
    def pp(self):
        for i in range(self.H):
            for j in range(self.W):
                if self.obs[i][j]>=0:
                    print(self.obs[i][j], end=' ')
                else:
                    print('*', end=' ')
            print()
        
    def input(self):
        while True:
            i, j = input('请输入坐标:').split()
            _, done = self.act(int(i), int(j))
            if done:
                print('game over!!!')
                print(self.real_state)
                print(self.state_type)
                break
            self.pp()
        
# 测试用
# env=Env(5, 5, 5)
# env.input()

强化学习算法library库：(集成库)

https://github.com/Denys88/rl_games

https://github.com/Domattee/gymTouch

个人github博客地址：
https://devilmaycry812839668.github.io/