data-raw/income.R

In rwright88/counties: US county data

library(readr)
library(dplyr)
library(stringr)
library(purrr)

url <- "https://www2.census.gov/programs-surveys/saipe/datasets/"

files <- str_c(url, c(
  "2000/2000-state-and-county/est00all.dat",
  "2001/2001-state-and-county/est01all.dat",
  "2002/2002-state-and-county/est02all.dat",
  "2003/2003-state-and-county/est03all.dat",
  "2004/2004-state-and-county/est04all.txt",
  "2005/2005-state-and-county/est05all.txt",
  "2006/2006-state-and-county/est06all.txt",
  "2007/2007-state-and-county/est07all.txt",
  "2008/2008-state-and-county/est08all.txt",
  "2009/2009-state-and-county/est09all.txt",
  "2010/2010-state-and-county/est10all.txt",
  "2011/2011-state-and-county/est11all.txt",
  "2012/2012-state-and-county/est12all.txt",
  "2013/2013-state-and-county/est13all.txt",
  "2014/2014-state-and-county/est14all.txt",
  "2015/2015-state-and-county/est15all.txt",
  "2016/2016-state-and-county/est16all.txt"
))

read_income <- function(file) {
  # name of file must contain year of data
  read_fwf(
    file,
    col_positions = fwf_cols(
      state = c(1, 2),
      county = c(4, 6),
      income_hh = c(134, 139)
    ),
    col_types = "cci",
    na = c("NA", "", ".")
  ) %>%
    mutate(year = as.integer(str_extract(file, "\\d{4}")))
}

income <- map_dfr(files, read_income) %>%
  mutate(county_fips = str_c(
    str_pad(state, 2, pad = "0"),
    str_pad(county, 3, pad = "0"))
  ) %>%
  filter(str_sub(county_fips, 3, 5) != "000") %>%
  select(county_fips, year, income_hh) %>%
  arrange(county_fips, year)

devtools::use_data(income, overwrite = TRUE)