5 ms·

I was going to point out that you can use R and tigris to get it, but that is only Zip Code Tabulation Areas, which don't map one-to-one to zip codes. Guess th

I was going to point out that you can use R and tigris to get it, but that is only Zip Code Tabulation Areas, which don't map one-to-one to zip codes. Guess the Census Bureau and the USPS just can't get along.

maxerickson 9y ago
USPS defines zipcodes as lists of addresses rather than areas. They can be synthesized into areas, but they aren't designed as areas or intended to be useful for spatial analysis.
- abakker 9y ago
  if you have actual addresses, I'd mention this API - https://locate.pitneybowes.com/geocode https://locate.pitneybowes.com/geocode - which is good for turning addresses into lat/long coordinates, which are probably better for spatial analysis. Of course, with map data, its always nice to actually just look at the data points in aggregate first.
kevin_thibedeau 9y ago
They should have been using census tracts all along. They're small enough for epidemiology in urban areas.