incorrectly calculating the 538 urbanization index for each census tract
https://twitter.com/NateSilver538/status/1200498385125433344
You can install urbanindex from github with:
install.packages("edavidaja/urbanindex")
if that doesn’t work–have you met
{renv}
?
library(urbanindex)
head(pop_within_5_mi)
#> GEOID pop_within_five population
#> 1 01001020100 58291 1923
#> 2 01001020200 58555 2028
#> 3 01001020300 64819 3476
#> 4 01001020400 68921 3831
#> 5 01001020500 71947 9883
#> 6 01001020600 50997 3705
a differently wrong approach uses st_is_within_distance
to identify
tract centroids within five miles of each other, available on the
centroid-self-join
branch