count_white = 2.5
for i in range(self.size):
for j in range(self.size):
if board[i][j] == 1:
count_black += 1
elif board[i][j] == 2:
count_white += 1
if piece_type == 1:
diff = count_black - count_white
else:
diff = count_white - count_black
return diff
if __name__ == "__main__":
N = 5
go_game = Game(N)
game_piece_type, previous_board, current_board, go_game.n_move = go_game.read_input(
)
go_game.set_board(game_piece_type, previous_board, current_board)
player = Minimax()
if go_game.new_game:
player.cache = {}
open("cache.txt", "w").close()
player.side = game_piece_type
next_action = player.get_input(go_game, game_piece_type)
go_game.n_move += 2
go_game.write_output(next_action)
reward = LOSS
max_q_value = -1.0
first_iteration = True
self.states_to_update_opponent.reverse()
for state, move in self.states_to_update_opponent:
if self.opponent == 1:
base_state_action_q, orientation, base_state = self.state_q_values_X(go, state)
else:
base_state_action_q, orientation, base_state = self.state_q_values_O(go, state)
if first_iteration:
base_state_action_q[move] = reward
first_iteration = False
else:
base_state_action_q[move] = base_state_action_q[move] \
+ self.alpha * (self.gamma * max_q_value - base_state_action_q[move])
max_q_value = max(base_state_action_q.values())
self.opponent_actions = []
self.states_to_update_opponent = []
if __name__ == "__main__":
N = 5
go_game = Game(N)
game_piece_type, previous_board, board = go_game.read_input()
go_game.set_board(game_piece_type, previous_board, board)
player = Q_learning_agent()
Q_learning_agent.identity = game_piece_type
player.fight(0)
next_action = player.get_input(go_game, game_piece_type)
go_game.write_output(next_action)