Advanced PTC imputer

[cczhu]

A spin-off from these two comments in #32:

• determine how many more locations x years could be made into permanent count location-years if we started relaxing the criteria to be a permanent station. If there are tons of locations, or tons of years in current permanent count locations, that would be considered permanent with a small relaxing of criteria, it would justify further work on imputation.
• create a slightly more advanced imputer that handles a small number of missing months using Bagheri's algorithm (eg. estimating AADT by using conversion factors from the nearest year).

We should also explore using elements of Bagheri's algorithm to handle PTC data imputation. Check if this leads to more accurate growth factors, but that seems like its own issue too.