def alpha_beta(current_board, current_player, remaining_depth, alpha, beta,
maximizing_player):
if remaining_depth == 0 or is_terminal_node(current_board, current_player):
return eval_board(current_board, current_player)
if maximizing_player:
v = min_eval_board
for y in range(board_size):
for x in range(board_size):
legal, _ = is_legal_move(current_board, x, y, current_player)
if legal:
board_temp, total_piece_taken = execute_move(
copy.deepcopy(current_board), x, y, current_player)
v = max(
v,
alpha_beta(board_temp, current_player,
remaining_depth - 1, alpha, beta, False))
alpha = max(alpha, v)
if beta <= alpha:
break
return v
else: # minimizingPlayer
v = max_eval_board
for y in range(board_size):
for x in range(board_size):
legal, _ = is_legal_move(current_board, x, y, current_player)
if legal:
board_temp, total_piece_taken = execute_move(
copy.deepcopy(current_board), x, y, current_player)
v = min(
v,
alpha_beta(board_temp, current_player,
remaining_depth - 1, alpha, beta, True))
beta = min(beta, v)
if beta <= alpha:
break
return v
def is_legal_move(current_board, x, y, player):
if not (0 <= x < board_size and 0 <= y < board_size):
return False, 'Out of range'
if current_board[y][x] != '0':
return False, 'Box is filled'
_, total_piece_taken = execute_move(copy.deepcopy(current_board), x, y,
player)
if total_piece_taken == 0:
return False, 'No piece taken!'
return True, 'OK'
def minimax(current_board, current_player, remaining_depth, maximizing_player):
if remaining_depth == 0 or is_terminal_node(current_board, current_player):
return eval_board(current_board, current_player)
legal_moves = get_list_legal_moves(current_board, current_player)
if maximizing_player:
best_value = min_eval_board
for x, y in legal_moves:
board_temp, _ = execute_move(copy.deepcopy(current_board), x, y,
current_player)
v = minimax(board_temp, current_player, remaining_depth - 1, False)
best_value = max(best_value, v)
else: # minimizing_player
best_value = max_eval_board
for x, y in legal_moves:
board_temp, _ = execute_move(copy.deepcopy(current_board), x, y,
current_player)
v = minimax(board_temp, current_player, remaining_depth - 1, True)
best_value = min(best_value, v)
return best_value
def best_move(current_board, current_player, cpu_mode, deepest_depth):
points, max_points = 0, 0
x_move_to, y_move_to = -1, -1
legal_moves = get_list_legal_moves(current_board, current_player)
for x, y in legal_moves:
board_temp, _ = execute_move(copy.deepcopy(current_board), x, y,
current_player)
if cpu_mode == 1:
points = minimax(board_temp, current_player, deepest_depth, True)
elif cpu_mode == 2:
points = alpha_beta(current_board, current_player, deepest_depth,
min_eval_board, max_eval_board, True)
elif cpu_mode == 3:
return random_bot(current_board, current_player)
if points > max_points:
max_points = points
x_move_to, y_move_to = x, y
return x_move_to, y_move_to
print()
print_board()
print('Player cannot play! Game ended!')
print('Score User: '******'1')))
print('Score AI : ' + str(eval_board(board, '2')))
sys.exit()
else:
print('No moves, Player ' + player + 'skipped')
continue
elif player == '1': # 1st AI's turn
x_move_to, y_move_to = best_move(board, player, cpu_mode1,
deepest_depth)
if (x_move_to, y_move_to) != (-1, -1):
board, total_piece_taken = execute_move(
board, x_move_to, y_move_to, player)
print('1st AI played (X Y): ' + str(x_move_to) + ' ' +
str(y_move_to))
print('# of pieces taken: ' + str(total_piece_taken))
else: # 2nd AI's turn
x_move_to, y_move_to = best_move(board, player, cpu_mode2,
deepest_depth)
if (x_move_to, y_move_to) != (-1, -1):
board, total_piece_taken = execute_move(
board, x_move_to, y_move_to, player)
print('2nd AI played (X Y): ' + str(x_move_to) + ' ' +
str(y_move_to))
print('# of pieces taken: ' + str(total_piece_taken))
else: