self.model.step(action)
if self.iteration_count >= 200:
terminal = True
else:
terminal = star.terminal(self.model.state)
reward = star.reward(self.model.state, terminal)
brain_state = star.state(self.model.state)
return (brain_state, reward, terminal)
if __name__ == "__main__":
config = bonsai_ai.Config(sys.argv)
brain = bonsai_ai.Brain(config)
model = CartPole()
sim = CartpoleSimulator(brain, 'CartpoleSimulator', config)
sim.model = model
render = None
if '--render' in sys.argv:
log.info('rendering')
from render import Viewer
render = True
viewer = Viewer()
viewer.model = model
log.info('starting simulation...')
def simulate(self, action):
print("simulate")
state = {
"test_number": 0,
"test_double": 1,
"test_float": 2,
"test_int64": 3,
"test_int32": 4,
"test_uint32": 5,
"test_uint32": 6,
#"test_bool": True,
#"test_string": 0,
}
terminal = self.iteration_count > 5
reward = 1.0
return (state, reward, terminal)
def episode_finish(self):
print("episode_finish")
sys.exit()
if __name__ == "__main__":
config = bonsai_ai.Config()
brain = bonsai_ai.Brain(config)
sim = BasicSimulator(brain, "the_simulator")
while sim.run():
continue
return {
"currentState": state,
"comfort": self.simulation.comfort,
"lastAction": self.simulation.lastAction
}
def _shape_reward(self):
"""
Return a reward for approaching the target. Max 1, min -2.
"""
return 0
def reward_shaped(self):
"""Reward for approaching target"""
return self.simulation.get_subjects_comfort()
def _get_numeric_state_from(self, state):
if state not in self._statesMap:
self._statesMap[state] = len(self._statesMap)
return self._statesMap[state]
if __name__ == "__main__":
logger.info('starting...')
config = bsa.Config(sys.argv)
brain = bsa.Brain(config)
sim = InteractiveHomeBridge(brain, "home_sim")
while sim.run():
continue