You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Right now when I run an episode, it does not stop and just runs forever, since the following function is not implemented:
def _check_episode_done(self, list_state):
# ======== to implement ========
return False
Couple of questions:
1- What episode's stopping condition did you use in the paper?
2- How does the environment uses the traffic-flow file? Does it go to the first of the traffic-flow file at every time that I hit env.reset()?
I am asking this, since if I force the episode to stop at a given simulation time (the output of env.get_current_time()), the returned state by env.reset() is always a vector of zeros.
The text was updated successfully, but these errors were encountered:
Stopping condition is the total simulation time, specified in arg.cnt .
This correlates with the CityFlow simulator. Env.reset() will basically reload the traffic flow file (or reset the system state, including the simulation time and the traffic flow)
I see that it controls dic_exp_conf["RUN_COUNTS"] in the generator function, but still the env does not return done=True. Does it harm if I modify def _check_episode_done(self, list_state):?
Right now when I run an episode, it does not stop and just runs forever, since the following function is not implemented:
Couple of questions:
1- What episode's stopping condition did you use in the paper?
2- How does the environment uses the traffic-flow file? Does it go to the first of the traffic-flow file at every time that I hit
env.reset()
?I am asking this, since if I force the episode to stop at a given simulation time (the output of
env.get_current_time()
), the returned state byenv.reset()
is always a vector of zeros.The text was updated successfully, but these errors were encountered: