Joint Tour Participation - Optimize Two Person Households

[danielsclint]

trafficstars

Disclaimer: This is feature enhancement request. If I understand the code correctly, this should save some time while not impacting the results of the model. If my understanding of the code is incorrect, disregard. This came out of investigating #288.

If a two person household has been chosen to participate in a joint tour, it doesn't seem necessary to send the two person household participants through the Monte Carlo simulation. If the two-person household has already been determined to participate in a joint tour, the Monte Carlo simulation is going to iterate until both participants have been chosen. It would seem easier to process two person household in a deterministic manner and only use the Monte Carlo simulation to determine participants in households with three or more people.

https://github.com/ActivitySim/activitysim/blob/7b57c94e12d27a23f86c8988117e885ae754c44e/activitysim/abm/models/joint_tour_participation.py#L100-L201

[dhensle]

#569