Skip to content
TheChatArchiveThe Chat Archive
gpt-3.5-turbo-0301· OpenAIAi2 WildChat-4.8M (AI2)

根据这一段要求""" Recursively search values of all succeeding nodes, taking maximum of children…

  1. User

    根据这一段要求""" Recursively search values of all succeeding nodes, taking maximum of children when current player is the agent (self.player) and minimum for opponent.

            Parameters:
                s: the current state
    
            Return:
                Tuple(value, action): the node value and the best action (if exists)
            
            Note: before you perform an action, you might need to copy the original state for in-place update.
            """在                # TODO的地方填入代码           def minimax_search(s: State) -> Tuple:
    

    end, winner = s.game_end() value, action = None, None if end: if winner == -1: value = 0 else: value = (1 if winner == self.player else -1) else: # TODO pass return value, action

        return minimax_search(state)[1]
    
  2. Assistant

    def minimax_search(s: State) -> Tuple: end, winner = s.game_end() value, action = None, None if end: if winner == -1: value = 0 else: value = (1 if winner == self.player else -1) else: if s.player == self.player: value = -float('inf') for a in s.available_actions(): new_s = s.copy().apply_action(a) child_value, _ = minimax_search(new_s) if child_value > value: value = child_value action = a else: value = float('inf') for a in s.available_actions(): new_s = s.copy().apply_action(a) child_value, _ = minimax_search(new_s) if child_value < value: value = child_value action = a return value, action

        return minimax_search(state)[1]
    
  3. User

    你没有缩进

  4. Assistant

    抱歉,可能因为复制时缺少了缩进,这里为你再次给出缩进后的代码:

    def minimax_search(s: State) -> Tuple: end, winner = s.game_end() value, action = None, None if end: if winner == -1: value = 0 else: value = (1 if winner == self.player else -1) else: if s.player == self.player: value = -float('inf') for a in s.available_actions(): new_s = s.copy().apply_action(a) child_value, _ = minimax_search(new_s) if child_value > value: value = child_value action = a else: value = float('inf') for a in s.available_actions(): new_s = s.copy().apply_action(a) child_value, _ = minimax_search(new_s) if child_value < value: value = child_value action = a return value, action

    return minimax_search(state)[1]

根据这一段要求""" Recursively search values of all succeeding… · The Chat Archive