This is a utility function for performing BISG. It operates on a dataframe obtained from the Census Bureau via the provided eiCompare helper function.
compute_p_g(counts, cols = c("whi", "bla", "his", "asi", "oth"))
A vector containing the probability of selecting a person from each geographic unit.
A tibble containing counts (divided amongst constituent groups) per geographic units (rows).
The columns denoting the constituent groups within each geographic units.