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
{{ message }}
This repository has been archived by the owner on Jan 17, 2020. It is now read-only.
Something that I can think of right now is basically to fetch all stop_times that are available and put them into a sorted-set like https://github.com/clojure/data.avl
Then sort those based on departure_time or on stop to have very fast query time.
NOTE: since clojure uses shared data structures, only the overhead of maintaining the indexes of the sets is added
Original issue taken from #174
Right now we rely solely on the trips available at the requested DateTime to find out the necessary stop.times every time that we try to find a trip.
https://github.com/hiposfer/kamal/blob/master/src/hiposfer/kamal/services/routing/directions.clj#L207
However, having the trips means that we can also know the stop.times and the stops available.
I think this would speed up the routing since we are doing all of these computations anyway but without this level of preprocessing.
I am not sure how much would this speed things up since we dont have a gtfs benchmark. So for the time being this is rather a nice to have.
The text was updated successfully, but these errors were encountered: