Revise travel day and region variable

[sgreenbury]

Closes #86.

This PR continues from closed #82 rebased on main with new branch name.

This PR:

• Adds configurable common day of travel for households and revise the filtering of NTS trips to before matching to increase matches with non-missing trips. When config is not common travel day, then any day is chosen from the set of configured nts_days_of_week
• Serializes a chosen_day randomly picked during feasible zone assignment
• Revises the subset of commuting flows to exclude WFH as it seems NTS would not record a trip
• Relatedly, the PR adds "scaling" of flows that can be configured currently using the part_time_work_prob parameter (see below). The related census question 49 ("Where do you mainly work?") suggests that on a randomly chosen work day an individual would not necessarily be commuting. A float for scaling can be applied to the flows applicable if use_percentages=False (see Discrepancy in no. of work trips: census vs nts/acbm #86 (comment) for additional discussion)
• Adds initial approach for matching for unmatched individuals after household matching
• Revises the variable used for region to "PSUStatsReg_B01ID" as this is included for 2021 and 2022
• Adds tests

[sgreenbury]

Currently config of scaling of the flows is implemented through the part_time_work_prob. The name of this parameter should be revised - however, this will change the ID of the config so it would be preferred to update the name in a subsequent PR once the current parameter testing is complete.