Geocode using TIGER address features — geocode

geocode_tiger() geocodes addr vectors using TIGER address range geometries downloaded from census.gov for a specific county.

The geocoding precision is the level of matching achieved for each address:

range: is the most precise; the address matched on street name and ZIP code, and the number was within one of the ranges; point locations are interpolated from the geometry of the best match using the street number
street: is less precise; the address matched on street name and ZIP code, but the number was not within one of the ranges; point locations are interpolated as the centroid of all matching range geometries
unmatched: no precision; the address did not match any street name and ZIP code combinations; point locations are <null feature> (which is equivalent to NA)

Usage

geocode_tiger(x, county, year, offset = 0)

Arguments

x: an addr vector
county: character string of county identifier
year: character year of tigris product
offset: number of meters to offset geocode from street line

Value

a s2_geography vector of geocoded point locations; an attribute ("geocode_precision") is added as a factor with levels range, street, or unmatched.

Details

If a street number-name has multiple in range and parity matches, then the best match is chosen based on the smallest width of ranges and then on the range with the closest midpoint.

Examples

if (FALSE) { # \dontrun{
gcd <-
  geocode_tiger(as_addr(voter_addresses()[1:100]),
                county = "39061", year = "2024", offset = 20)

head(gcd)

table(attr(gcd, "geocode_precision"))

# consider addresses matched to street but out of range missing
gcd[attr(gcd, "geocode_precision") == "street"] <- NA_character_

# convert to s2_cell
s2::as_s2_cell(gcd)

# create leaflet map
leaflet::leaflet(wk::wk_coords(gcd)) |>
  leaflet::addTiles() |>
  leaflet::addCircleMarkers( lng = ~x, lat = ~y, label = ~ feature_id)
} # }