def step(self, action_e):
# TODO implement clock_speed: step only if self.clock.to_step()
if self.done:
return self.reset()
action_e = util.nanflatten(action_e)
env_info_dict = self.u_env.step(action_e)
reward_e, state_e, done_e = self.env_space.aeb_space.init_data_s(ENV_DATA_NAMES, e=self.e)
for (a, b), body in util.ndenumerate_nonan(self.body_e):
env_info_a = self.get_env_info(env_info_dict, a)
reward_e[(a, b)] = env_info_a.rewards[b]
state_e[(a, b)] = env_info_a.states[b]
done_e[(a, b)] = env_info_a.local_done[b]
self.done = (util.nonan_all(done_e) or self.clock.get('t') > self.max_timestep)
return reward_e, state_e, done_e
def step(self, action_e):
assert len(action_e) == 1, 'OpenAI Gym supports only single body'
# TODO implement clock_speed: step only if self.clock.to_step()
if self.done: # t will actually be 0
return self.reset()
action = action_e[(0, 0)]
(state, reward, done, _info) = self.u_env.step(action)
if util.get_lab_mode() == 'dev':
self.u_env.render()
reward_e, state_e, done_e = self.env_space.aeb_space.init_data_s(ENV_DATA_NAMES, e=self.e)
for (a, b), body in util.ndenumerate_nonan(self.body_e):
reward_e[(a, b)] = reward
state_e[(a, b)] = state
done_e[(a, b)] = done
self.done = (util.nonan_all(done_e) or self.clock.get('t') > self.max_timestep)
return reward_e, state_e, done_e
def space_step(self, action_e):
# TODO implement clock_speed: step only if self.clock.to_step()
if self.done:
return self.space_reset()
action_e = util.nanflatten(action_e)
env_info_dict = self.u_env.step(action_e)
reward_e, state_e, done_e = self.env_space.aeb_space.init_data_s(
ENV_DATA_NAMES, e=self.e)
for (a, b), body in util.ndenumerate_nonan(self.body_e):
env_info_a = self._get_env_info(env_info_dict, a)
reward_e[(a, b)] = env_info_a.rewards[b] * self.reward_scale
state_e[(a, b)] = env_info_a.states[b]
done_e[(a, b)] = env_info_a.local_done[b]
self.done = (util.nonan_all(done_e) or self.clock.t > self.max_t)
logger.debug(
f'Env {self.e} step reward_e: {reward_e}, state_e: {state_e}, done_e: {done_e}'
)
return reward_e, state_e, done_e
def test_nonan_all(v, isall):
assert util.nonan_all(v) == isall
def test_nonan_all(v, isall):
assert util.nonan_all(v) == isall
def __bool__(self):
return util.nonan_all(self.data)
def __bool__(self):
return util.nonan_all(self.data)
