Open stefancoe opened 1 year ago
periods: [20, 0, 5, 9, 15, 18] labels: ['EA', 'AM', 'MD', 'PM', 'EV']
I don't think this will work because the periods
list needs to be monotonically increasing, otherwise you will get a pandas error on this line complaining "ValueError: bins must increase monotonically".
periods: [0, 5, 9, 15, 18, 20, 24] labels: ['EA', 'AM', 'MD', 'PM', 'EV', 'EA']
Having EA
at the end of labels
could be problematic. The mandatory tour scheduling model uses the order of labels to create in_period
and out_period
combination when calculating logsum. See the trace file below, notice how AM-EA, MD-EA, PM-EA, and EV-EA combinations are created. This could be what you want if you are modeling overnight travel. But note that if you do not update the tour_departure_and_duration_segments.csv, which is used to look up the start and end time for logsum calculation, you will end up having negative duration. (Edit: duration
is naively calculated in the ActivitySim source code using end - start, which can cause negative values.) You probably want to double check to make sure this is not happening.
We are using the skim time period label 'EA' twice in our network los configuration. Is this skim/time period configuration supported? If not, how should it be configured without having to create two sets of (the same) skims with different labels?
https://github.com/psrc/psrc_activitysim/blob/main/configs_dev/network_los.yaml#L24
**Edit: Perhaps this is the proper way to do it: periods: [20, 0, 5, 9, 15, 18] labels: ['EA', 'AM', 'MD', 'PM', 'EV']