open_ai_gym_environment.py 文件源码

def get_training_data_page(self, num_samples):
        """
        Returns a TrainingDataPage with shuffled, transformed transitions from
        replay memory.

        :param num_samples: Number of transitions to sample from replay memory.
        """
        states, actions, rewards, next_states, next_actions, terminals,\
            possible_next_actions = self.sample_memories(num_samples)
        return TrainingDataPage(
            np.array(states, dtype=np.float32),
            np.array(actions, dtype=np.float32),
            np.array(rewards, dtype=np.float32),
            np.array(next_states, dtype=np.float32),
            np.array(next_actions, dtype=np.float32),
            np.array(possible_next_actions, dtype=np.float32),
            None, None, np.logical_not(terminals, dtype=np.bool)
        )