Table 1: Locations lumped by the main location area in the code.
Table 2: By location code.
Seeing the detailed location codes illustate that there is some noise in the dataset, for example, the “bd” code is a temp record forbo rrow direct. But rather than get bogged down in detail, what is offered here is a starting point.
We could also easily add more detailed groups, by year or month, for example.