def step_wait(self) -> VecEnvStepReturn:
# wield the stored action for env.step
for idx, env in enumerate(self.envs):
obs, self.buf_rews[idx], self.buf_dones[idx], self.buf_infos[
idx] = env.step(self.actions[idx])
if any(self.buf_dones[idx].tolist()):
# save final observation where user can get it, then reset
self.buf_infos[idx]["terminal_observation"] = obs
obs = self.envs[idx].reset()
self._save_obs(idx, obs)
return (dict_to_obs(self.observation_space,
copy_obs_dict(self.buf_obs)),
np.copy(self.buf_rews), np.copy(self.buf_dones),
copy.deepcopy(self.buf_infos))
def _obs_from_buf(self) -> VecEnvObs:
return dict_to_obs(self.observation_space, copy_obs_dict(self.buf_obs))
def reset(self) -> VecEnvObs:
for env_idx, each_env in enumerate(self.envs):
obs = each_env.reset()
self._save_obs(env_idx, obs)
return dict_to_obs(self.observation_space, copy_obs_dict(self.buf_obs))