Grid-based accessibility computations

[rmkujala]

To really enable door-to-door accessibility estimates one may want to

The current algorithm implementations in gtfspy allow for door-to-door routing (via adding artificial stops), but the implementation probably is not efficient enough for computing travel times matrices in large grids (see e.g. http://mapple.fi).

To solve this problem, one could try out the following approach:

1. Discover Pareto-optimal stop-to-stop journeys without any walking egress or access legs.
2. Create candidate journeys between the grid-based origin (O) and grid-based destination (D) by combing (walk(O->i) + journey(i->j) + walk(j->D) ), where i and j are labels for stops.
3. Gather all candidate journeys, where i is within walking distance from O, and j is within walking distance from D, together into collection C.
4. Compute P, the set of Pareto-optimal journeys in C.
5. Compute travel time and transfer estimates using the journeys of P.

[sevamoo]

Thanks a lot for open sourcing your great works here. I was wondering if there was any implementation regarding this suggestion so far?
I read your paper and codes here and was wondering if any part of it could be used for this issue?
I have a similar problem, where I want to calculate building to building travel/boarding times in Switzerland.
Appreciate a lot any comments.
Best
Vahid