I have timings data for a route like so:
(route is from A to D)
Stretch | Headway (mm:ss)
A to B | 03:57
B to C | 02:38
C to D | 05:16
So this is calculated / estimated time published by the agency to make it more digestible for the passengers. It's not like the route has any hard breaks in the between.
Some trips start from A and go all the way to D.
One trip may start from A and terminate at C, then passengers have to disembark and wait for the next train to get till D.
Yet another trip may start from B and go to D,
And yet another may only cover B to C stretch.
It's obvious that (C to D) is just half the frequency of (B to C), meaning every other train halts at C instead of going all the way to D.
But I'm not really getting the (A to B) relation with (B to C).
There is a fixed first trip and last trip for the whole route, and we have timings per stop, meaning the both the first and last trains of the day do the full A to D stretch.
So how would one express this in GTFS? Might it be a better idea to just multiply it all into distinct trips and then go shaving trips off to get to the stated frequency?
And related but may not be applicable in this case : Are overlapping frequencies possible in frequencies.txt?