def f():
if request.method == 'POST':
data = request.get_json()
print(data)
state = []
me_moves = []
you_moves = []
your_move = None
for i in range(256):
state.append(int(data['Inputs']['input1'][0][str(i + 1)]))
if int(data['Inputs']['input1'][0][str(i + 1)]) == 1:
me_moves.append((i // 16, i % 16))
if int(data['Inputs']['input1'][0][str(i + 1)]) == -1:
you_moves.append((i // 16, i % 16))
if len(me_moves) == 0:
game = GameState()
origin = None
if len(you_moves) > 0: your_move = you_moves[0]
if len(you_moves) > 1: return 'Error: you_moves > 1'
else:
game = pickle.load(open('game.pkl', 'rb'))
origin = pickle.load(open('origin.pkl', 'rb'))
pre_state = pickle.load(open('state.pkl', 'rb'))
your_move = None
for i in range(16):
for j in range(16):
if state[i * 16 + j] != pre_state[i * 16 + j] and state[
i * 16 + j] == -1:
if your_move != None:
return 'Error: your_move not none'
your_move = (i, j)
if your_move != None:
print('Moved', your_move)
game.move(*your_move)
player_icon = None
if len(me_moves) == len(you_moves):
player_icon = 'X'
elif len(me_moves) == (len(you_moves) - 1):
player_icon = 'O'
else:
return 'Error: no player icon'
recom_moves = predict(state, model_name='gomoku_nas.hdf5')
print(recom_moves)
print(origin)
print(game)
tar, tree = MCTSPolicy(player=player_icon).move(
game, recom_moves, 100, origin)
game.move(*tar)
print(game)
pickle.dump(game, open('game.pkl', 'wb'))
pickle.dump(tree, open('origin.pkl', 'wb'))
state[tar[0] * 16 + tar[1]] = 1
pickle.dump(state, open('state.pkl', 'wb'))
return str(tar[0] * 16 + tar[1])
from gameplay import play_game
from policies import RandomPolicy, MCTSPolicy
import numpy as np
import networkx as nx
player_policies = [MCTSPolicy(), RandomPolicy()]
# For reproducibility
np.random.seed(0)
games = []
for i in range(100):
games.append(play_game(player_policies))
graphs = [game[0] for game in games]
dot_graph_combined = nx.compose_all(graphs)
dot_graph = nx.to_pydot(dot_graph_combined)
dot_graph.set_graph_defaults(fontname='Courier')
dot_graph.write_png('multiple_game_graph.png')
def __init__(self, players, turn_id, board):
self.board = board
## to track of whose turn is now
self.players = players
self.turn_id = turn_id
self.status = Game.GAME_STATUS[0]
self.turn = self.players[self.turn_id]
self.flag_for_drawing_canvas = False
# in case a move needs to be made through Random
self.random_policy = RandomPolicy()
## MCTSPolicy(a, b) -- a is player, b is for an opponent
self.mctsObj_X = MCTSPolicy(self.players[0],
self.players[1],
board=self.board)
self.mctsObj_O = MCTSPolicy(self.players[1],
self.players[0],
board=self.board)
self.mctsObjs = [self.mctsObj_X, self.mctsObj_O]
"""
model_dir = "./analysis-tools/models_ex/"
model_w_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924weights.h5"
model_json_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924in_json.json"
"""
self.model_based_policy = None # ModelPolicy(model_dir, model_w_file, model_json_file)
for each_player in self.players:
print(each_player.get_policy_mode())
if each_player.get_policy_mode() == "MODEL":
model_dir = "./analysis-tools/models_ex/"
#model_w_file = "model_2020-01-09-17-23-04_BEST_SO_FAR_WITH_Early_Stop-0.730-upto2-0.925weights.h5"
#model_json_file ="model_2020-01-09-17-23-04_BEST_SO_FAR_WITH_Early_Stop-0.730-upto2-0.925in_json.json"
# Second best
#model_w_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924weights.h5"
#model_json_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924in_json.json"
# very good Best
#model_w_file = "model_2020-01-11-11-16-48_win_sample_focus_0.875_weights.h5"
#model_json_file = "model_2020-01-11-11-16-48_win_sample_focus_0.875_in_json.json"
# third good
#model_json_file = "model_2020-01-11-20-11-26_win_sample_focus_0.91_in_json.json"
#model_w_file = "model_2020-01-11-20-11-26_win_sample_focus_0.91_weights.h5"
## LOOKS best so far.. waiting to be done
model_json_file = "model_2020-01-11-20-39-46_winAndLoss_sample_focus_0.71_in_json.json"
model_w_file = "model_2020-01-11-20-39-46_winAndLoss_sample_focus_0.71_weights.h5"
# Done
model_json_file = "model_2020-01-11-21-07-12_winAndLoss_sample_focus_0.75_in_json.json"
model_w_file = "model_2020-01-11-21-07-12_winAndLoss_sample_focus_0.75_weights.h5"
#
model_json_file = "model_2020-01-12-08-47-16_winAndLoss_sample_focus_0.749_in_json.json"
model_w_file = "model_2020-01-12-08-47-16_winAndLoss_sample_focus_0.749_weights.h5"
#model_w_file ="model_2020-01-12-18-59-53_winAndLoss_sample_focus_0.71_weights.h5"
#model_json_file = "model_2020-01-12-18-59-53_winAndLoss_sample_focus_0.71_in_json.json"
# Done -- 1K weighted for preventing lose
#model_w_file = "model_2020-01-12-19-29-34_winAndLoss_Loss1KWeights_sample_focus_0.70_weights.h5"
#model_json_file = "model_2020-01-12-19-29-34_winAndLoss_Loss1KWeights_sample_focus_0.70_in_json.json"
# Done
model_w_file = "model_2020-01-12-21-02-40_winAndLoss_sample_focus_0.733_weights.h5"
model_json_file = "model_2020-01-12-21-02-40_winAndLoss_sample_focus_0.733_in_json.json"
# DOne -- below are from a buggy weighting scheme..
model_json_file = "model_2020-01-12-21-40-17_winAndLoss_combinedWithUniq_sample_focus_0.649_in_json.json"
model_w_file = "model_2020-01-12-21-40-17_winAndLoss_combinedWithUniq_sample_focus_0.649_weights.h5"
# WORST SO FAR
model_w_file = "model_2020-01-13-07-27-36_winAndLoss_combinedWithUniq_sample_focus_0.41_weights.h5"
model_json_file = "model_2020-01-13-07-27-36_winAndLoss_combinedWithUniq_sample_focus_0.41_in_json.json"
# BEST SO FAR
model_json_file = "model_2020-01-13-21-02-40_winAndLoss_Loss1KWeights_sample_focus_0.718_in_json.json"
model_w_file = "model_2020-01-13-21-02-40_winAndLoss_Loss1KWeights_sample_focus_0.718_weights.h5"
# DONE
model_json_file = "model_2020-01-14-00-19-22_winAndLoss_combinedWithUniq_sample_focus_0.65_in_json.json"
model_w_file = "model_2020-01-14-00-19-22_winAndLoss_combinedWithUniq_sample_focus_0.65_weights.h5"
# DONE
#model_json_file = "model_2020-01-14-21-34-33_winAndLoss_withOneHotEncodeForLabel_sample_focus_0.7108_in_json.json"
#model_w_file = "model_2020-01-14-21-34-33_winAndLoss_withOneHotEncodeForLabel_sample_focus_0.7108_weights.h5"
model_obj = each_player.get_model_obj()
#self.model_based_policy = ModelPolicy(model_obj) #model_dir, model_w_file, model_json_file)
break
self.game_id = uuid.uuid1()
class Game():
GAME_STATUS = ["Playing", "End", "Draw"]
def __init__(self, players, turn_id, board):
self.board = board
## to track of whose turn is now
self.players = players
self.turn_id = turn_id
self.status = Game.GAME_STATUS[0]
self.turn = self.players[self.turn_id]
self.flag_for_drawing_canvas = False
# in case a move needs to be made through Random
self.random_policy = RandomPolicy()
## MCTSPolicy(a, b) -- a is player, b is for an opponent
self.mctsObj_X = MCTSPolicy(self.players[0],
self.players[1],
board=self.board)
self.mctsObj_O = MCTSPolicy(self.players[1],
self.players[0],
board=self.board)
self.mctsObjs = [self.mctsObj_X, self.mctsObj_O]
"""
model_dir = "./analysis-tools/models_ex/"
model_w_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924weights.h5"
model_json_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924in_json.json"
"""
self.model_based_policy = None # ModelPolicy(model_dir, model_w_file, model_json_file)
for each_player in self.players:
print(each_player.get_policy_mode())
if each_player.get_policy_mode() == "MODEL":
model_dir = "./analysis-tools/models_ex/"
#model_w_file = "model_2020-01-09-17-23-04_BEST_SO_FAR_WITH_Early_Stop-0.730-upto2-0.925weights.h5"
#model_json_file ="model_2020-01-09-17-23-04_BEST_SO_FAR_WITH_Early_Stop-0.730-upto2-0.925in_json.json"
# Second best
#model_w_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924weights.h5"
#model_json_file = "model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924in_json.json"
# very good Best
#model_w_file = "model_2020-01-11-11-16-48_win_sample_focus_0.875_weights.h5"
#model_json_file = "model_2020-01-11-11-16-48_win_sample_focus_0.875_in_json.json"
# third good
#model_json_file = "model_2020-01-11-20-11-26_win_sample_focus_0.91_in_json.json"
#model_w_file = "model_2020-01-11-20-11-26_win_sample_focus_0.91_weights.h5"
## LOOKS best so far.. waiting to be done
model_json_file = "model_2020-01-11-20-39-46_winAndLoss_sample_focus_0.71_in_json.json"
model_w_file = "model_2020-01-11-20-39-46_winAndLoss_sample_focus_0.71_weights.h5"
# Done
model_json_file = "model_2020-01-11-21-07-12_winAndLoss_sample_focus_0.75_in_json.json"
model_w_file = "model_2020-01-11-21-07-12_winAndLoss_sample_focus_0.75_weights.h5"
#
model_json_file = "model_2020-01-12-08-47-16_winAndLoss_sample_focus_0.749_in_json.json"
model_w_file = "model_2020-01-12-08-47-16_winAndLoss_sample_focus_0.749_weights.h5"
#model_w_file ="model_2020-01-12-18-59-53_winAndLoss_sample_focus_0.71_weights.h5"
#model_json_file = "model_2020-01-12-18-59-53_winAndLoss_sample_focus_0.71_in_json.json"
# Done -- 1K weighted for preventing lose
#model_w_file = "model_2020-01-12-19-29-34_winAndLoss_Loss1KWeights_sample_focus_0.70_weights.h5"
#model_json_file = "model_2020-01-12-19-29-34_winAndLoss_Loss1KWeights_sample_focus_0.70_in_json.json"
# Done
model_w_file = "model_2020-01-12-21-02-40_winAndLoss_sample_focus_0.733_weights.h5"
model_json_file = "model_2020-01-12-21-02-40_winAndLoss_sample_focus_0.733_in_json.json"
# DOne -- below are from a buggy weighting scheme..
model_json_file = "model_2020-01-12-21-40-17_winAndLoss_combinedWithUniq_sample_focus_0.649_in_json.json"
model_w_file = "model_2020-01-12-21-40-17_winAndLoss_combinedWithUniq_sample_focus_0.649_weights.h5"
# WORST SO FAR
model_w_file = "model_2020-01-13-07-27-36_winAndLoss_combinedWithUniq_sample_focus_0.41_weights.h5"
model_json_file = "model_2020-01-13-07-27-36_winAndLoss_combinedWithUniq_sample_focus_0.41_in_json.json"
# BEST SO FAR
model_json_file = "model_2020-01-13-21-02-40_winAndLoss_Loss1KWeights_sample_focus_0.718_in_json.json"
model_w_file = "model_2020-01-13-21-02-40_winAndLoss_Loss1KWeights_sample_focus_0.718_weights.h5"
# DONE
model_json_file = "model_2020-01-14-00-19-22_winAndLoss_combinedWithUniq_sample_focus_0.65_in_json.json"
model_w_file = "model_2020-01-14-00-19-22_winAndLoss_combinedWithUniq_sample_focus_0.65_weights.h5"
# DONE
#model_json_file = "model_2020-01-14-21-34-33_winAndLoss_withOneHotEncodeForLabel_sample_focus_0.7108_in_json.json"
#model_w_file = "model_2020-01-14-21-34-33_winAndLoss_withOneHotEncodeForLabel_sample_focus_0.7108_weights.h5"
model_obj = each_player.get_model_obj()
#self.model_based_policy = ModelPolicy(model_obj) #model_dir, model_w_file, model_json_file)
break
self.game_id = uuid.uuid1()
def show_progress_on_canvas(self, a_boolean_flag):
self.flag_for_drawing_canvas = a_boolean_flag
def set_to_next_player(self):
next_turn = (self.turn_id + 1) % len(self.players)
self.turn_id = next_turn
self.turn = self.players[self.turn_id]
def is_end(self):
if self.is_draw():
return True
for each_player in self.players:
if self.check_end_status(each_player):
return True
return False
def check_end_status(self, a_player):
if self.board.is_win(a_player):
return True
return False
def get_input(self):
prompt = "%s 's Turn\n" % (self.turn)
input_from_user = input(prompt)
r_c_in_list = input_from_user.split("_")
r, c = r_c_in_list[0], r_c_in_list[1]
r_int = ord(r) - ord('a')
c_int = int(c)
return r_int, c_int
def validate_input(self):
available_pos = self.board.get_available_positions()
while True:
r, c = self.get_input()
if available_pos.get((r, c), 0) == 1:
break
print("Try again. Your input")
return r, c
# def convert_sequence_moves_to_vector(self):
# individual_sequence = [0] * 9
# for item in self.board.sequences_of_movements:
# turn_for_this_move = item.get("turn")
# move_made_for_this_move = item.get("position")
# individual_sequence[move_made_for_this_move - 1] = 1 if turn_for_this_move == "X" else 2
#
# return np.array([individual_sequence])
def play_game(self):
turn_id = 0
game_log = {
'game_uuid': self.get_game_id(),
'play_modes': {
'X': self.players[0].get_policy_mode(),
'O': self.players[1].get_policy_mode()
},
'board_size': self.board.row,
'winner': "",
'sequence': {}
}
canvas_for_drawing = None
if self.flag_for_drawing_canvas:
#turtle.setup(500, 500)
canvas_for_drawing = Draw()
is_draw_gametie = False
"""
# Below block must be gone
# from model_loader import ModelBasedAgent
# model_dir = "./analysis-tools/models_ex/"
# model_w_file = model_dir + "current_best.h5" #"model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924weights.h5"
# model_json_file = model_dir + "current_best.json" #model_2020-01-09-15-15-06_BEST_SO_FAR_WITH_Early_Stop-0.90-upto2-0.924in_json.json"
# model_agent_obj = ModelBasedAgent(model_w_file, model_json_file)
# mlModel = model_agent_obj.get_model()
"""
while self.check_end_status(self.turn) != True:
print(self.board)
test_instance = self.board.convert_sequence_moves_to_vector()
#print(test_instance)
if self.turn.get_player_type() == Player.PTYPE_HUMAN:
# TODO -- this part is just to make a simplified interface of modelbased movement
# later, this will be the part of Policy as a ModelPolicy class
# for now, we assume player O would be model.. as X is always starting first
#test_instance = np.array([an_instance])
#prediction_move = mlModel.predict_proba(test_instance)[0]
#pp = model_agent_obj.predict_proba(test_instance)[0]
#UT.print_three_arrays(test_instance[0], pp, prediction_move)
#move_by_prediction = np.argmax(pp) + 1
#r_e, c_e = self.board.indices_to_coordinate(move_by_prediction)
#print("R:%d C:%d \t i_e:%d R_e:%d C_e:%d" % (r_v, c_v, move_by_prediction, r_e, c_e))
r_v, c_v = self.validate_input()
else: # when Player is an agent
if self.turn.get_policy_mode() == "MODEL":
model_structure = 3 # 0 for regular, 1 for two tower, 2 for conv2d, 3 for conv2d+twoTowers
r_v, c_v = self.turn.model_based_policy.move(
self.board, model_structure)
elif self.turn.get_policy_mode() == "MCTS":
if self.turn.get_marker() == "O":
r_v, c_v = self.mctsObj_O.move(self.board)
# TODO -- this part is just to make a simplified interface of modelbased movement
# This could be a place for ModelBased action
elif self.turn.get_marker() == "X":
if self.turn.get_policy_mode() == "RANDOM":
self.random_policy = RandomPolicy()
r_v, c_v = self.random_policy.move(self.board)
# print("AM I HERE FOR RANDOM")
else:
r_v, c_v = self.mctsObj_X.move(self.board)
elif self.turn.get_policy_mode() == "RANDOM":
self.random_policy = RandomPolicy()
r_v, c_v = self.random_policy.move(self.board)
self.board.set_a_move(r_v, c_v, self.turn)
UT.print_as_log(self.board.get_available_positions())
## Drawing on canvas
if self.flag_for_drawing_canvas:
canvas_for_drawing.move_and_draw(r_v, c_v,
self.turn.get_marker())
if self.check_end_status(self.turn):
print("FinalResult: %s" % (self.turn.get_marker()))
print(self.board)
print(self.board.convert_sequence_moves_to_vector())
#UT.print_as_log("Winning and so ending this game")
UT.print_as_log(self.board.sequences_of_movements)
game_log['winner'] = self.turn.get_marker()
game_log['sequence'] = self.board.sequences_of_movements
break
elif self.is_draw():
is_draw_gametie = True
print("FinalResult: Draw")
#UT.print_as_log("Draw.... so, exiting the game")
print(self.board)
print(self.board.convert_sequence_moves_to_vector())
game_log['winner'] = "D"
game_log['sequence'] = self.board.sequences_of_movements
break
else:
self.set_to_next_player()
## for writing a message to the canvas
if self.flag_for_drawing_canvas:
result_message = "Game result -- Winner is %s" % (
game_log.get("winner"))
if is_draw_gametie:
result_message = "Game result : Draw"
canvas_for_drawing.write_text(result_message)
canvas_for_drawing.exit_on_click()
#canvas_for_drawing.reset_canvas()
#turtle.TurtleScreen._RUNNING = True
json_str = game_log #json.dumps(game_log)
return json_str
def a_move_for_agent(self):
r, c = self.a_move_for_agent_helper()
return r, c
## this is the function for an agent to come up with a smarter decision
def a_move_for_agent_helper(self):
all_available_positions_dict = self.board.get_available_positions()
random_move_index = np.random.randint(
0, len(all_available_positions_dict), 1)[0]
r, c = list(all_available_positions_dict.keys())[random_move_index]
return r, c
def is_draw(self):
if len(self.board.get_available_positions()) < 1:
return True
return False
@staticmethod
def load_a_game(afile):
move_sequences = UT.read_a_game(afile)
if move_sequences:
Game.parse_history(move_sequences)
@staticmethod
def parse_history(adict, message_str=None):
winner = adict.get("winner", None)
if winner == None:
print("Something is wrong")
sys.exit(1)
move_sequences = adict.get("sequence", None)
turtle.hideturtle()
board_obj_from_history = Board(3, 3, 3)
# below obj is for drawing the board on a canvas.
# if you don't like, you can make it comment
draw_board_obj = Draw()
for each_move in move_sequences:
player_marker = each_move.get("turn")
r_index, c_index = each_move.get("xy")
p = Player("test", player_marker, 1)
board_obj_from_history.set_a_move(r_index, c_index, p)
draw_board_obj.move_and_draw(r_index, c_index, player_marker)
print(board_obj_from_history)
draw_board_obj.write_text(
("Winner is: %s -- sampled %s" % (winner, str(message_str))))
time.sleep(3)
draw_board_obj.turtle_obj.getpen().clear()
draw_board_obj.turtle_obj.getscreen().clearscreen()
# draw_board_obj.exit_on_click()
# or
def get_game_id(self):
return str(self.game_id)
"""
Plays many games and then plots the cumulative win rates of the players.
The player policies can be chosen from MCTS and Random.
"""
from gameplay import play_game
from policies import RandomPolicy, MCTSPolicy
from visualization import visualize_mcts_tree
import networkx as nx
import numpy as np
# Choose the player policies here:
MCTS_vs_Random = [MCTSPolicy(player='X'), RandomPolicy()]
Random_vs_MCTS = [RandomPolicy(), MCTSPolicy(player='O')]
MCTS_vs_MCTS = [MCTSPolicy(player='X'), MCTSPolicy(player='O')]
Random_vs_Random = [RandomPolicy(), RandomPolicy()]
experiments = [[MCTSPolicy(player='X'), RandomPolicy()],
[MCTSPolicy(player='X'), RandomPolicy()],
[RandomPolicy(), MCTSPolicy(player='O')],
[RandomPolicy(), MCTSPolicy(player='O')],
[MCTSPolicy(player='X'),
MCTSPolicy(player='O')],
[MCTSPolicy(player='X'),
MCTSPolicy(player='O')], [RandomPolicy(),
RandomPolicy()],
[RandomPolicy(), RandomPolicy()]]
names = [
'x_mcts_vs_o_random_1', 'x_mcts_vs_o_random_2', 'x_random_vs_o_mcts_1',
def play_game(player_policies):
game = GameState()
origin = None
# Inform the player policies that a new game is starting (so they can reset their current move pointers)
for player_policy in player_policies:
if type(player_policy) is MCTSPolicy:
player_policy.reset_game()
while game.winner() is None:
for player_policy in player_policies:
# print("\n================ ( It is {}'s move. ) ================".format(game.turn()))
# Player 1 : AI
if game.turn() is "X":
# player_policy.__init__('X')
# recommend moves : [(0, 0), (0, 1)] etc.
recom_moves = [(0, 0), (0, 1), (0, 2), (0, 3), (0, 4), (0, 5), (0, 6), (0, 7), (0, 8), (0, 9), (0, 10), (0, 11), (0, 12), (0, 13), (0, 14), (0, 15),
(1, 0), (1, 1), (1, 2), (1, 3), (1, 4), (1, 5), (1, 6), (1, 7), (1, 8), (1, 9), (1, 10), (1, 11), (1, 12), (1, 13), (1, 14), (1, 15),
(2, 0), (2, 1), (2, 2), (2, 3), (2, 4), (2, 5), (2, 6), (2, 7), (2, 8), (2, 9), (2, 10), (2, 11), (2, 12), (2, 13), (2, 14), (2, 15),
(3, 0), (3, 1), (3, 2), (3, 3), (3, 4), (3, 5), (3, 6), (3, 7), (3, 8), (3, 9), (3, 10), (3, 11), (3, 12), (3, 13), (3, 14), (3, 15),
(4, 0), (4, 1), (4, 2), (4, 3), (4, 4), (4, 5), (4, 6), (4, 7), (4, 8), (4, 9), (4, 10), (4, 11), (4, 12), (4, 13), (4, 14), (4, 15),
(5, 0), (5, 1), (5, 2), (5, 3), (5, 4), (5, 5), (5, 6), (5, 7), (5, 8), (5, 9), (5, 10), (5, 11), (5, 12), (5, 13), (5, 14), (5, 15),
(6, 0), (6, 1), (6, 2), (6, 3), (6, 4), (6, 5), (6, 6), (6, 7), (6, 8), (6, 9), (6, 10), (6, 11), (6, 12), (6, 13), (6, 14), (6, 15),
(7, 0), (7, 1), (7, 2), (7, 3), (7, 4), (7, 5), (7, 6), (7, 7), (7, 8), (7, 9), (7, 10), (7, 11), (7, 12), (7, 13), (7, 14), (7, 15),
(8, 0), (8, 1), (8, 2), (8, 3), (8, 4), (8, 5), (8, 6), (8, 7), (8, 8), (8, 9), (8, 10), (8, 11), (8, 12), (8, 13), (8, 14), (8, 15),
(9, 0), (9, 1), (9, 2), (9, 3), (9, 4), (9, 5), (9, 6), (9, 7), (9, 8), (9, 9), (9, 10), (9, 11), (9, 12), (9, 13), (9, 14), (9, 15),
(10, 0), (10, 1), (10, 2), (10, 3), (10, 4), (10, 5), (10, 6), (10, 7), (10, 8), (10, 9), (10, 10), (10, 11), (10, 12), (10, 13), (10, 14), (10, 15),
(11, 0), (11, 1), (11, 2), (11, 3), (11, 4), (11, 5), (11, 6), (11, 7), (11, 8), (11, 9), (11, 10), (11, 11), (11, 12), (11, 13), (11, 14), (11, 15),
(12, 0), (12, 1), (12, 2), (12, 3), (12, 4), (12, 5), (12, 6), (12, 7), (12, 8), (12, 9), (12, 10), (12, 11), (12, 12), (12, 13), (12, 14), (12, 15),
(13, 0), (13, 1), (13, 2), (13, 3), (13, 4), (13, 5), (13, 6), (13, 7), (13, 8), (13, 9), (13, 10), (13, 11), (13, 12), (13, 13), (13, 14), (13, 15),
(14, 0), (14, 1), (14, 2), (14, 3), (14, 4), (14, 5), (14, 6), (14, 7), (14, 8), (14, 9), (14, 10), (14, 11), (14, 12), (14, 13), (14, 14), (14, 15),
(15, 0), (15, 1), (15, 2), (15, 3), (15, 4), (15, 5), (15, 6), (15, 7), (15, 8), (15, 9), (15, 10), (15, 11), (15, 12), (15, 13), (15, 14), (15, 15)]
"""
recom_moves = []
while True:
try:
m, n = input("recommend : ").split()
m = int(m)
n = int(n)
tar = (m, n)
recom_moves.append(tar)
except:
break
print(recom_moves)
"""
# tar, tree = player_policy.move(game, recom_moves, 100, origin)
tar, tree = MCTSPolicy(player='X').move(game, recom_moves, 100, origin)
origin = copy.deepcopy(tree)
print("MCTS: " + str(tar))
# Player 2 : Human
else:
m, n = input("which Do you want? : ").split()
m = int(m)
n = int(n)
tar = (m, n)
game.move(*tar)
print(game)
if game.winner() is not None:
break
return game.winner()
except:
break
print(recom_moves)
"""
# tar, tree = player_policy.move(game, recom_moves, 100, origin)
tar, tree = MCTSPolicy(player='X').move(game, recom_moves, 100, origin)
origin = copy.deepcopy(tree)
print("MCTS: " + str(tar))
# Player 2 : Human
else:
m, n = input("which Do you want? : ").split()
m = int(m)
n = int(n)
tar = (m, n)
game.move(*tar)
print(game)
if game.winner() is not None:
break
return game.winner()
# main
winner = play_game([MCTSPolicy(player='X'), MCTSPolicy(player='O')])
print('Game over. Winner is Player ' + str(winner))