censable creates a common interface to tidycensus, with memoized data functions. With a growing redistricting universe in R, it seems right to partition out some use cases, to allow imports to be more targeted and to stop packages from repeating each other.

Installation

You can download the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("christopherkenny/censable")

Example

The primary data functions are build_dec and build_acs

de <- build_dec(geography = 'county', state = 'DE')
#> Getting data from the 2010 decennial Census
#> Downloading feature geometry from the Census website.  To cache shapefiles for use in future sessions, set `options(tigris_use_cache = TRUE)`.
#> Using Census Summary File 1

The build_* functions take arguments on geography and state, at a minimum. If you’re looking at a smaller piece of a state, you can use county to subset it. Year defaults to 2010, but will be updated to 2020 around October 2021, following the API release of the 2020 Census data.

The build_* functions are a convenient wrapper around tidycensus, with an eye towards collecting data on total population and voting age population by race. Total population variables begin with the prefix pop, while voting age population variables begin with the prefix vap.

Additionally, there are memoised versions of the build_* functions:

de <- mem_build_dec(geography = 'county', state = 'DE')
#> Getting data from the 2010 decennial Census
#> Downloading feature geometry from the Census website.  To cache shapefiles for use in future sessions, set `options(tigris_use_cache = TRUE)`.
#> Using Census Summary File 1

These speed up repeated calls, allowing you to have code that is self contained and reproducable, without taking the time hit of having to redownload the data each time.