Measure types check

[wanghalan]

I had a conversation with Joanna today about types in data, and apparently, there are a few distinctions we can make. Specifically, we can disambiguate the difference between data types and units. For example, an integer can be a data type, but dollars can be a unit.

There are also a few pieces of additional information we parsed today:

1. Location where existing data types are used: https://github.com/uva-bi-sdad/community/blob/2a4d0a88de1ecf0d65b4c39b7c8995d7e4325b4f/dev/community.js#L3201-L3216

2. Set of values in measure_info files:

{‘’, ‘count’, ‘minutes’, ‘density’, ‘dollar’, ‘dollars’, ‘percent’, ‘rate’, ‘internet speed’, ‘MOE’, ‘cost’, ‘integer’, ‘per 1k’, ‘rate per 100K’, ‘drive time’, None, ‘float’, ‘per 100k’, ‘index’}

3. Set of values in measure_type columns in data:

{‘index’, ‘median’, ‘prevalence estimate’, ‘percentage’, ‘percent’, ‘MOE’, ‘minutes’, ‘per 100k’, ‘integer’, ‘float’, ‘mean’, ‘quotient’, 
‘per_1k’, ‘cost’, ‘density’, ‘rate per 100k’, ‘count’, ‘dollars’, ‘proportion’, ‘rate per 100K’}

Their symmetric difference is:

{'', 'prevalence estimate', 'rate per 100k', 'rate', 'percentage', 'per_1k', 'drive time', 'mean', 'quotient', 'internet speed', 'per 1k', None, 'dollar', 'median', 'proportion'}

We anticipate some standards on this can help elucidate how to properly check for data types, data ranges, and will trickle down to measure_info structure and test routines

[wanghalan]

Been having conversations about where this type of metadata should exist in measure_info files or in the csv.xz, so development ongoing