class BotAgent:
def __init__(self, env):
"""An agent based on a GOFAI bot."""
self.env = env
self.on_reset()
def on_reset(self):
self.bot = Bot(self.env)
def act(self, obs=None, update_internal_state=True, *args, **kwargs):
action = self.bot.get_action()
if update_internal_state:
self.bot.take_action(action)
return {'action': action}
def analyze_feedback(self, reward, done):
pass
# To make things simple, only allow random left/right/fwd moves, and opening of doors
if action in (mission.actions.left,
mission.actions.right,
mission.actions.forward):
break
fwd_pos = mission.agent_pos + mission.dir_vec
fwd_cell = mission.grid.get(*fwd_pos)
if action == mission.actions.toggle and fwd_cell is not None and fwd_cell.type == 'door':
break
before_optimal_actions.append(action)
else:
optimal_actions.append(action)
expert.take_action(action)
obs, reward, done, info = mission.step(action)
total_reward += reward
episode_steps += 1
if done:
if reward > 0:
num_success += 1
total_steps += episode_steps
if reward <= 0:
assert episode_steps == mission.max_steps # Is there another reason for this to happen ?
if options.verbose:
print('FAILURE on %s, seed %d, reward %.2f' %
(level_name, mission_seed, reward))
break
