Make it clear that residential -> other accommodation conversations are labeled as other accommodations. This means that if you're doing analysis on changes in only residential housing, if you exclude other accommodations, you are excluding some of the residential housing unit reductions. You would need to filter by both occ_prop and occ_init.
Clarify that units_net is only calculated when 1) both units_init and u_prop contain real values or 2) if a CofO has been issued and u_prop has a value. This means that in other situations when not enough data is available, some units are being excluded from any calculations that use units_net.
Suggest that users also exclude suspended and disapproved records from analysis
Disclose 80% = complete assumption
Data dictionary
Add floodplain fields to data dictionary
Process explanation
Update checklist and summaries to reflect the revised sequence of scripts
Limitations:
Data dictionary
Process explanation