def __init__(self,
size,
player1='human_player',
player2='human_player',
mode=2,
temp=0.01):
super().__init__()
self.g = GomokuBoard((size[0], size[1])) # 初始化游戏内容
self.H = size[0]
self.W = size[1]
self.init_ui() # 初始化游戏界面
self.last_pos = (-1, -1)
self.res = 0 # 记录那边获得了胜利
self.operate_status = 0 # 游戏操作状态。0为游戏中（可操作），1为游戏结束闪烁过程中（不可操作
self.player1 = player1
self.player2 = player2
self.temp = temp
self.mode = mode
def game_restart(self, res):
"""游戏出现开始"""
self.res = res # 标记谁获胜了
# 1. 显示游戏结束的信息
if self.res == 1:
# QMessageBox.about(self, '游戏结束', 'Player1获胜!')
inform = QMessageBox.information(self, '游戏结束', '黑棋获胜',
QMessageBox.Yes, QMessageBox.Yes)
elif self.res == 2:
# QMessageBox.about(self, '游戏结束', 'Player2获胜!')
inform = QMessageBox.information(self, '游戏结束', '白棋获胜',
QMessageBox.Yes, QMessageBox.Yes)
elif self.res == 3:
# QMessageBox.about(self, '游戏结束', '平局!')
inform = QMessageBox.information(self, '游戏结束', '平局!',
QMessageBox.Yes, QMessageBox.Yes)
else:
raise ValueError('当前游戏结束的标志位为' + self.res +
'. 而游戏结束的标志位必须为1, 2 或 3')
# 2. 游戏重新开始的操作
self.res = 0
self.operate_status = 0
self.g = GomokuBoard((self.H, self.W)) # 重新初始化游戏内容
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50) # 重新绘制游戏界面
def __init__(self, size=(8, 8), init_model=None):
# params of the board and the game
print(size)
self.board_width = size[1]
self.board_height = size[0]
self.board = GomokuBoard(size=(self.board_width, self.board_height))
self.game = GomokuGame(self.board)
# training params
self.learn_rate = 2e-3
self.lr_multiplier = 1.0 # adaptively adjust the learning rate based on KL
self.temp = 1.0 # the temperature param
self.n_playout = 400 # num of simulations for each move
self.c_puct = 5
self.buffer_size = 10000
self.batch_size = 512 # mini-batch size for training
self.data_buffer = deque(maxlen=self.buffer_size)
self.play_batch_size = 1
self.epochs = 5 # num of train_steps for each update
self.kl_targ = 0.02
self.check_freq = 50
self.game_batch_num = 3000
self.best_win_ratio = 0.0
self.all_loss = []
# num of simulations used for the pure mcts, which is used as
# the opponent to evaluate the trained policy
self.pure_mcts_playout_num = 1000
if init_model:
# start training from an initial policy-value net
self.policy_value_net = PolicyValueNet(self.board_width,
self.board_height,
model_file=init_model)
else:
# start training from a new policy-value net
self.policy_value_net = PolicyValueNet(self.board_width,
self.board_height)
self.mcts_player = MCTSPlayer(self.policy_value_net.policy_value_fn,
c_puct=self.c_puct,
n_playout=self.n_playout,
is_selfplay=1)
class GomokuWindow(QMainWindow):
def __init__(self,
size,
player1='human_player',
player2='human_player',
mode=2,
temp=0.01):
super().__init__()
self.g = GomokuBoard((size[0], size[1])) # 初始化游戏内容
self.H = size[0]
self.W = size[1]
self.init_ui() # 初始化游戏界面
self.last_pos = (-1, -1)
self.res = 0 # 记录那边获得了胜利
self.operate_status = 0 # 游戏操作状态。0为游戏中（可操作），1为游戏结束闪烁过程中（不可操作
self.player1 = player1
self.player2 = player2
self.temp = temp
self.mode = mode
def init_ui(self):
# """初始化游戏界面"""
# # 1. 确定游戏界面的标题，大小和背景颜色
self.setObjectName('MainWindow')
self.setWindowTitle('AlphaZero Gomoku with Forbidden rules')
self.setFixedSize(50 + self.H * 40, 50 + self.W * 40)
title = QLabel(self)
title.setText('AlphaZero五子棋')
title.move((self.H) * 17, 0)
author = QLabel(self)
author.setText('Fan Gaoyige')
author.move((self.H - 1) * 40, self.W * 40)
id = QLabel(self)
id.setText('1728405039')
id.move((self.H - 1) * 40, 20 + self.W * 40)
#print(50+self.H*40 ,50+self.W*40)
# self.setStyleSheet('#MainWindow{background-color: green}')
palette = QPalette()
# palette.setBrush(QPalette.Window, QBrush(QPixmap('img/board.jpg')))
palette.setBrush(QPalette.Background, QBrush(QPixmap("img/board.png")))
self.setPalette(palette)
# 2. 开启鼠标位置的追踪。并在鼠标位置移动时，使用特殊符号标记当前的位置
self.setMouseTracking(True)
# 3. 鼠标位置移动时，对鼠标位置的特殊标记
self.corner_widget = CornerWidget(self)
self.corner_widget.repaint()
self.corner_widget.hide()
self.show()
@run_with_exc
def paintEvent(self, e):
# """绘制游戏内容"""
def draw_map():
# """绘制棋盘"""
qp.setPen(QPen(QColor(0, 0, 0), 2, Qt.SolidLine)) # 棋盘的颜色为黑色
# # 绘制横线
for x in range(self.H):
qp.drawLine(40 * (x + 1), 40, 40 * (x + 1), self.H * 40)
# # 绘制竖线
for y in range(self.W):
qp.drawLine(40, 40 * (y + 1), self.W * 40, 40 * (y + 1))
# # 绘制棋盘中的黑点
# qp.setBrush(QColor(0, 0, 0))
# key_points = [(4, 4), (12, 4), (4, 12), (12, 12), (8, 8)]
# for t in key_points:
# qp.drawEllipse(QPoint(40 * t[0], 40 * t[1]), 5, 5)
def draw_pieces():
# """绘制棋子"""
# # 绘制黑棋子
qp.setPen(QPen(QColor(0, 0, 0), 1, Qt.SolidLine))
# qp.setBrush(QColor(0, 0, 0))
for x in range(self.H):
for y in range(self.W):
if self.g.chessboard[x, y] == 1:
# if self.flash_cnt % 2 == 1 and (x, y) in self.flash_pieces:
# continue
radial = QRadialGradient(40 * (x + 1), 40 * (y + 1),
15, 40 * x + 35,
40 * y + 35) # 棋子的渐变效果
radial.setColorAt(0, QColor(96, 96, 96))
radial.setColorAt(1, QColor(0, 0, 0))
qp.setBrush(QBrush(radial))
qp.drawEllipse(QPoint(40 * (x + 1), 40 * (y + 1)), 15,
15)
# # 绘制白棋子
qp.setPen(QPen(QColor(160, 160, 160), 1, Qt.SolidLine))
# qp.setBrush(QColor(255, 255, 255))
for x in range(self.H):
for y in range(self.W):
if self.g.chessboard[x, y] == 2:
# if self.flash_cnt % 2 == 1 and (x, y) in self.flash_pieces:
# continue
radial = QRadialGradient(40 * (x + 1), 40 * (y + 1),
15, 40 * x + 35,
40 * y + 35) # 棋子的渐变效果
radial.setColorAt(0, QColor(255, 255, 255))
radial.setColorAt(1, QColor(160, 160, 160))
qp.setBrush(QBrush(radial))
qp.drawEllipse(QPoint(40 * (x + 1), 40 * (y + 1)), 15,
15)
if hasattr(self, 'g'): # 游戏还没开始的话，就不用画了
qp = QPainter()
qp.begin(self)
draw_map() # 绘制棋盘
draw_pieces() # 绘制棋子
qp.end()
def keyPressEvent(self, e):
if e.key() == Qt.Key_Space and self.mode == 2:
turn = self.player1
while self.mode == 2:
move = turn.get_action(self.g, self.temp)
self.g.do_move(move)
result = self.g.get_result()
if result == 0:
turn = self.player2 if turn == self.player1 else self.player1
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50)
else:
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50)
self.game_restart(result)
break
@run_with_exc
def mouseMoveEvent(self, e):
if self.mode == 2:
return
# 1. 首先判断鼠标位置对应棋盘中的哪一个格子
mouse_x = e.windowPos().x()
mouse_y = e.windowPos().y()
if 25 <= mouse_x <= 615 and 25 <= mouse_y <= 615 and (
mouse_x % 40 <= 15
or mouse_x % 40 >= 25) and (mouse_y % 40 <= 15
or mouse_y % 40 >= 25):
game_x = int((mouse_x + 15) // 40) - 1
game_y = int((mouse_y + 15) // 40) - 1
else: # 鼠标当前的位置不对应任何一个游戏格子，将其标记为(01, 01
game_x = -1
game_y = -1
# 2. 然后判断鼠标位置较前一时刻是否发生了变化
pos_change = False # 标记鼠标位置是否发生了变化
if game_x != self.last_pos[0] or game_y != self.last_pos[1]:
pos_change = True
self.last_pos = (game_x, game_y)
# 3. 最后根据鼠标位置的变化，绘制特殊标记
if pos_change and game_x != -1:
self.setCursor(Qt.PointingHandCursor)
if pos_change and game_x == -1:
self.setCursor(Qt.ArrowCursor)
if pos_change and game_x != -1:
self.corner_widget.move(25 + game_x * 40, 25 + game_y * 40)
self.corner_widget.show()
if pos_change and game_x == -1:
self.corner_widget.hide()
@run_with_exc
def mousePressEvent(self, e):
if self.mode == 2:
return
if not (hasattr(self, 'operate_status') and self.operate_status == 0):
return
if e.button() == Qt.LeftButton:
# 1. 首先判断按下了哪个格子
mouse_x = e.windowPos().x()
mouse_y = e.windowPos().y()
if (mouse_x % 40 <= 10
or mouse_x % 40 >= 30) and (mouse_y % 40 <= 10
or mouse_y % 40 >= 30):
game_x = int((mouse_x + self.H) // 40) - 1
game_y = int((mouse_y + self.W) // 40) - 1
else: # 鼠标点击的位置不正确
return
self.g.do_move((game_x, game_y))
# self.show()
# 2. 根据操作结果进行一轮游戏循环
res = self.g.get_result() # 判断游戏结果
if res != 0: # 如果游戏结果为“已经结束”，则显示游戏内容，并退出主循环
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50)
self.game_restart(res)
return
self.repaint(0, 0, self.H * 40 + 50,
self.W * 40 + 50) # 在游戏还没有结束的情况下，显示游戏内容，并继续下一轮循环
if self.mode == 1:
move = self.player2.get_action(self.g, self.temp)
self.g.do_move(move)
res = self.g.get_result() # 判断游戏结果
if res != 0: # 如果游戏结果为“已经结束”，则显示游戏内容，并退出主循环
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50)
self.game_restart(res)
return
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50)
def game_restart(self, res):
"""游戏出现开始"""
self.res = res # 标记谁获胜了
# 1. 显示游戏结束的信息
if self.res == 1:
# QMessageBox.about(self, '游戏结束', 'Player1获胜!')
inform = QMessageBox.information(self, '游戏结束', '黑棋获胜',
QMessageBox.Yes, QMessageBox.Yes)
elif self.res == 2:
# QMessageBox.about(self, '游戏结束', 'Player2获胜!')
inform = QMessageBox.information(self, '游戏结束', '白棋获胜',
QMessageBox.Yes, QMessageBox.Yes)
elif self.res == 3:
# QMessageBox.about(self, '游戏结束', '平局!')
inform = QMessageBox.information(self, '游戏结束', '平局!',
QMessageBox.Yes, QMessageBox.Yes)
else:
raise ValueError('当前游戏结束的标志位为' + self.res +
'. 而游戏结束的标志位必须为1, 2 或 3')
# 2. 游戏重新开始的操作
self.res = 0
self.operate_status = 0
self.g = GomokuBoard((self.H, self.W)) # 重新初始化游戏内容
self.repaint(0, 0, self.H * 40 + 50, self.W * 40 + 50) # 重新绘制游戏界面