Just like we allow for the user to leverage multiprocessing during route level processing within generate_edge_and_wait_values, we should do the same for stop time interpolation. Currently, for example, AC Transit takes about 95 seconds to handle linearly_interpolate_infill_times running on AC Transit's GTFS feed (out of a total runtime of 545 seconds with multiprocessing utilizing 4 cores).
Just like we allow for the user to leverage multiprocessing during route level processing within
generate_edge_and_wait_values, we should do the same for stop time interpolation. Currently, for example, AC Transit takes about 95 seconds to handlelinearly_interpolate_infill_timesrunning on AC Transit's GTFS feed (out of a total runtime of 545 seconds with multiprocessing utilizing 4 cores).