class Game:
    def __init__(self, h, w):
        # 行數(shù)
        self.h = h
        # 列數(shù)
        self.w = w
        # 棋盤
        self.L = [['-' for _ in range(w)] for _ in range(h)]
        # 當(dāng)前玩家 - 表示空 X先下 然后是O
        self.cur = 'X'
        # 游戲勝利者
        self.win_user = None
 
    # 檢查下完這步后有沒(méi)有贏 y是行 x是列 返回True表示贏
    def check_win(self, y, x):
        directions = [
            # 水平、垂直、兩個(gè)對(duì)角線方向
            (1, 0), (0, 1), (1, 1), (1, -1)
        ]
        player = self.L[y][x]
        for dy, dx in directions:
            count = 0
            # 檢查四個(gè)方向上的連續(xù)相同棋子
            for i in range(-4, 5):  # 檢查-4到4的范圍，因?yàn)槲遄舆B珠需要5個(gè)棋子
                ny, nx = y + i * dy, x + i * dx
                if 0 <= ny < self.h and 0 <= nx < self.w and self.L[ny][nx] == player:
                    count += 1
                    if count == 5:
                        return True
                else:
                    count = 0
        return False
 
    # 檢查能不能下這里 y行 x列 返回True表示能下
    def check(self, y, x):
        return self.L[y][x] == '-' and self.win_user is None
 
    # 打印棋盤 可視化用得到
    def __str__(self):
        # 確定行號(hào)和列號(hào)的寬度
        row_width = len(str(self.h - 1))
        col_width = len(str(self.w - 1))
        
        # 生成帶有行號(hào)和列號(hào)的棋盤字符串表示
        result = []
        # 添加列號(hào)標(biāo)題
        result.append(' ' * (row_width + 1) + ' '.join(f'{i:>{col_width}}' for i in range(self.w)))
        # 添加分隔線（可選）
        result.append(' ' * (row_width + 1) + '-' * (col_width * self.w))
        # 添加棋盤行
        for y, row in enumerate(self.L):
            # 添加行號(hào)
            result.append(f'{y:>{row_width}} ' + ' '.join(f'{cell:>{col_width}}' for cell in row))
        return '\n'.join(result)
 
    # 一步棋
    def set(self, y, x):
        if self.win_user or not self.check(y, x):
            return False
        self.L[y][x] = self.cur
        if self.check_win(y, x):
            self.win_user = self.cur
            return True
        self.cur = 'X' if self.cur == 'O' else 'O'
        return True
    #和棋
    def heqi(self):
        for y in range(self.h):
            for x in range(self.w):
                if self.L[y][x]=='-':
                    return False
        return True
    

#玩家自己下
def run_game01():
    g = Game(15, 15)
    while not g.win_user:
        # 打印當(dāng)前棋盤狀態(tài)
        while 1:
            print(g)
            try:
                y,x=input(g.cur+':').split(',')
                x=int(x)
                y=int(y)
                if g.set(y,x):
                    break
            except Exception as e:
                print(e)
    print(g)
    print('勝利者',g.win_user)
 
 
 

import torch
import torch.nn as nn
import torch.optim as optim
from game import Game

class MyMod(nn.Module):
    def __init__(self, input_channels=1, output_size=15*15):
        super(MyMod, self).__init__()
        
        # 定義卷積層，用于提取特征
        self.conv1 = nn.Conv2d(input_channels, 32, kernel_size=3, padding=1)  # 輸出 32 x 15 x 15
        self.conv2 = nn.Conv2d(32, 64, kernel_size=3, padding=1)  # 輸出 64 x 15 x 15
        self.conv3 = nn.Conv2d(64, 128, kernel_size=3, padding=1)  # 輸出 128 x 15 x 15
        
        # 定義全連接層，用于最后的得分預(yù)測(cè)
        self.fc1 = nn.Linear(128 * 15 * 15, 1024)  # 展平后傳入全連接層
        self.fc2 = nn.Linear(1024, output_size)  # 輸出 15*15 的得分預(yù)測(cè)

    def forward(self, x):
        # 卷積層 -> 激活函數(shù) -> 最大池化
        x = torch.relu(self.conv1(x))
        x = torch.relu(self.conv2(x))
        x = torch.relu(self.conv3(x))
        
        # 將卷積層輸出展平為一維
        x = x.view(x.size(0), -1)
        
        # 全連接層
        x = torch.relu(self.fc1(x))
        x = self.fc2(x)
        return x
    # 保存模型權(quán)重
    def save(self, path):
        torch.save(self.state_dict(), path)

    # 加載模型權(quán)重
    def load(self, path):
        self.load_state_dict(torch.load(path))

#改進(jìn)一下  output 把有棋子的地方的概率=0避免下這些地方
# 輸入Game對(duì)象和MyMod對(duì)象，用于得到概率最大的落棋點(diǎn) (行y, 列x)
def input_qi(g: Game, m: MyMod):
    # 獲取當(dāng)前棋盤狀態(tài)
    board_state = g.L  # 使用 game.L 獲取當(dāng)前棋盤的狀態(tài) (15x15的二維列表)
    
    # 將棋盤狀態(tài)轉(zhuǎn)換為PyTorch的Tensor并增加一個(gè)維度（batch_size = 1）
    board_tensor = torch.tensor([[1 if cell == 'X' else -1 if cell == 'O' else 0 for cell in row] for row in board_state], 
                                 dtype=torch.float32).unsqueeze(0).unsqueeze(0)  # 形狀變?yōu)?(1, 1, 15, 15)
    
    # 傳入模型獲取每個(gè)位置的得分
    output = m(board_tensor)
    
    # 將輸出轉(zhuǎn)為概率值（可以使用softmax來(lái)歸一化）
    probabilities = torch.softmax(output, dim=-1).view(g.h, g.w).detach().numpy()  # 變?yōu)?(15, 15) 大小
    
    # 將已有棋子的位置的概率設(shè)置為 -inf，避免選擇這些位置
    for y in range(g.h):
        for x in range(g.w):
            if board_state[y][x] != '-':
                probabilities[y, x] = -float('inf')  # 設(shè)置已經(jīng)有棋子的地方的概率為 -inf
    
    # 找到概率最大的落子點(diǎn)
    max_prob_pos = divmod(probabilities.argmax(), g.w)  # 得到最大概率的行列坐標(biāo)
    
    # 確保返回的是合法的位置
    y, x = max_prob_pos
    
    return (y, x), output  # 返回坐標(biāo)和模型輸出

import os
import torch
import torch.optim as optim
import torch.nn.functional as F
from mod import MyMod, input_qi, Game

# 兩個(gè)權(quán)重文件，分別代表 X 棋和 O 棋
MX = 'MX'
MO = 'MO'

# 加載模型，若文件不存在則初始化
def load_model(model, path):
    if os.path.exists(path):
        model.load(path)
        print(f"Loaded model from {path}")
    else:
        print(f"{path} not found, initializing new model.")
        # 這里可以加一些初始化模型的代碼，例如：
        # model.apply(init_weights) 如果需要初始化權(quán)重

# 初始化模型
modx = MyMod()
load_model(modx, MX)

modo = MyMod()
load_model(modo, MO)

# 定義優(yōu)化器
lr=0.001
optimizer_x = optim.Adam(modx.parameters(), lr=lr)
optimizer_o = optim.Adam(modo.parameters(), lr=lr)

# 損失函數(shù)：根據(jù)游戲結(jié)果調(diào)整損失
def compute_loss(winner: int, player: str, model_output):
    # 將目標(biāo)值轉(zhuǎn)換為相應(yīng)的張量
    if player == "X":
        if winner == 1:  # X 勝
            target = torch.tensor(1.0, dtype=torch.float32)
        elif winner == 0:  # 平局
            target = torch.tensor(0.5, dtype=torch.float32)
        else:  # X 輸
            target = torch.tensor(0.0, dtype=torch.float32)
    else:
        if winner == -1:  # O 勝
            target = torch.tensor(1.0, dtype=torch.float32)
        elif winner == 0:  # 平局
            target = torch.tensor(0.5, dtype=torch.float32)
        else:  # O 輸
            target = torch.tensor(0.0, dtype=torch.float32)

    # 確保目標(biāo)值的形狀和 model_output 一致，假設(shè) model_output 是單一的值
    target = target.unsqueeze(0).unsqueeze(0)  # 形狀變?yōu)?(1, 1)

    # 使用均方誤差損失計(jì)算
    return F.mse_loss(model_output, target)



# 訓(xùn)練模型的過(guò)程
def train_game():
    modx.train()
    modo.train()

    # 創(chuàng)建新的游戲?qū)嵗?
    game = Game(15, 15)  # 默認(rèn)是 15x15 棋盤
    # 反向傳播和優(yōu)化
    optimizer_x.zero_grad()
    optimizer_o.zero_grad()

    while not game.win_user:  # 游戲未結(jié)束
        # X 方落子
        x_move, x_output = input_qi(game, modx)  # 獲取落子位置和模型輸出（x_output 是模型的輸出）
        game.set(x_move[0], x_move[1])  # X 下棋
        
        if game.win_user:
            break
        
        # O 方落子
        
        o_move, o_output = input_qi(game, modo)  # 獲取落子位置和模型輸出（o_output 是模型的輸出）
        #print(o_move,game)
        game.set(o_move[0], o_move[1])  # O 下棋
       

    # 獲取比賽結(jié)果
    winner = 0 if game.heqi() else (1 if game.win_user == 'X' else -1)  # 1為X勝，-1為O勝，0為平局

    # 計(jì)算損失
    loss_x = compute_loss(winner, "X", x_output)  # 傳遞模型輸出給計(jì)算損失函數(shù)
    loss_o = compute_loss(winner, "O", o_output)  # 傳遞模型輸出給計(jì)算損失函數(shù)

    # 計(jì)算損失并進(jìn)行反向傳播
    loss_x.backward()
    loss_o.backward()

    # 更新權(quán)重
    optimizer_x.step()
    optimizer_o.step()
    print(game)
    return loss_x.item(), loss_o.item()


# 訓(xùn)練多個(gè)回合
def train(num_epochs,n):
    k=0
    for epoch in range(num_epochs):
        loss_x, loss_o = train_game()
        print(f"Epoch [{epoch+1}/{num_epochs}], Loss X: {loss_x}, Loss O: {loss_o}")
        k+=1
        if k==n:
            modo.save('MO')
            modx.save('MX')
            print('saved')
            k=0

# 開(kāi)始訓(xùn)練
train(50000,1000)

from game import Game
from mod import MyMod,input_qi


#玩家下X ai下O
def playX():
    m=MyMod()
    m.load('MO')
    g=Game(15,15)
    while 1:
        print(g)
        if g.heqi() or g.win_user:
            break
        while 1:
            try:
                r=input('X:')
                y,x=r.split(',')
                y=int(y)
                x=int(x)
                if g.set(y,x):
                    break
            except Exception as e:
                print(e)
        if g.heqi() or g.win_user:
            break
        while 1:
            (y,x),_=input_qi(g,m)
            if g.set(y,x):
                break
    print(g)
    print('winner',g.win_user)


#玩家下O ai下X
def playO():
    m=MyMod()
    m.load('MX')
    g=Game(15,15)
    while 1:
        
        if g.heqi() or g.win_user:
            break
        while 1:
            (y,x),_=input_qi(g,m)
            if g.set(y,x):
                break
        if g.heqi() or g.win_user:
            break
        print(g)
        while 1:
            try:
                r=input('O:')
                y,x=r.split(',')
                y=int(y)
                x=int(x)
                if g.set(y,x):
                    break
            except Exception as e:
                print(e)
    print(g)
    print('winner',g.win_user)

playX()