zippeR providers support for downloading geometric data
as sf objects. For five-digit ZCTAs, tigris is
used to access the TIGER/Line database. For three-digit ZCTAs, these are
downloaded from a
repository on GitHub maintained by the author. These are generalized
versions of the ZCTA geometry designed to provide a smaller file size
suitable for storage on GitHub. Nevertheless, each file is still
approximately 60MB in size. Users should carefully evaluate the
geometric data to ensure they are fit for purpose, and note that these
are provided without warranty.
One major difference between tigris::zctas() and
zipper::get_geometry() is that zippeR provides
support for automatically returning ZCTAs that intersect or whose
centroids lie within states and/or counties. This can be used with both
the five- and three-digit ZCTAs. In general, the intersect method will
yield more ZCTAs than are necessary while the centroid method will yield
too few. For example, for Missouri, we need to use the
includes argument to add several additional ZCTAs in whose
centroids do not fall within Missouri itself (they all lie in Iowa):
geo12 <- zi_get_geometry(year = 2012, state = "MO", method = "centroid",
includes = c("51640", "52542", "52573", "52626"))If you use the intersect method, the corresponding
excludes argument can be used to drop ZCTAs that are not
needed. Note that all ZCTAs passed to starts_with (which
accepts two-digit values), includes, and
excludes are validated to ensure they follow formatting
requirements. As noted above, the validation process does not ensure
that inputs correspond to valid ZCTAs at this time.
One additional option to note is that, when downloading nation-wide
ZCTA data, you can use the shift_geo argument to place
Alaska, Hawaii, and Puerto Rico in the lower left-hand corner of your
map:
Downloading the geometric data can be used for mapping, or simply to
produce a vector of all five- or three-digit ZCTAs in a given area.
These, in-turn, can be used to define areas where demographic data are
desired. The zi_get_demographics() wraps
tidycensus functions to get data using a single, uniform
API. For instance, we can download the total population for all ZCTAs
based on the 2010 Decennial Census:
Note: As of now, the Census Bureau has not released ZCTA data for the 2020 Decennial Census.
If you are working with output from zi_get_geometry(),
you can specify the GEOID vector in your call to limit the
number of ZCTAs returned: