def __init__(self, x_player, o_player):
self.players = [x_player, o_player]
self.board = Board()
self.player_number = 0
for player, piece in zip(self.players, [Board.X, Board.O]):
run_if_learner(player, lambda: player.reset())
player.set_board(self.board)
player.set_piece(piece)
def _train_batch(self, player1, player2, stat_type):
run_if_learner(self.player1, lambda: self.player1.enable_learning())
run_if_learner(self.player2, lambda: self.player2.enable_learning())
stats = self._init_stats()
for batch_number in range(self.num_batches):
winner = self._train_game(player1, player2)
stats[winner] += 1
self._show_stats(stat_type, stats)
return stats
def _train_game(self, player1, player2):
controller = GameController(player1, player2)
winner = None
while winner is None:
winner, _ = controller.make_move()
run_if_learner(player1, lambda: player1.store_state())
run_if_learner(player2, lambda: player2.store_state())
run_if_learner(player1, lambda: player1.set_reward(winner))
run_if_learner(player2, lambda: player2.set_reward(winner))
return winner
def test_run_if_learner_does_not_run_function_if_human(self):
run_if_learner(self.player, self.func)
self.func.assert_not_called()
def test_run_if_learner_runs_function_if_td_learner(self):
run_if_learner(self.player, self.func)
self.func.assert_called_once_with()
def get_and_load_player(self, player_type, piece):
player = player_types.get_player(player_type)
run_if_learner(player, lambda: player.load(piece))
return player
def swap_players(self, player1, player2):
run_if_learner(player1, lambda: player1.load(Board.O))
run_if_learner(player2, lambda: player2.load(Board.X))
return player2, player1