R/acs-variables.R

In tiernanmartin/NeighborhoodChangeTypology: Project: Neighborhood Change Typology for King County, WA

#' @title Prepare ACS Data Variables
#' @description Return a `tibble` of all of the American Community Survey data variables.
#' @param acs_data Tibble, description.
#' @param acs_tables desc
#' @param cpi desc
#' @param variable_template desc
#' @return a `tibble`

#' @rdname acs-variables
#' @export
make_acs_variables <- function(acs_data, acs_tables, cpi, variable_template){

  # PREPARE ACS DATA ROLES --------------------------------------------------------

  # Join each acs variable its respective indicator


  all_vars <- tidycensus::load_variables(2016, "acs5", cache = TRUE) %>%
    dplyr::transmute(SOURCE = "ACS",
                     VARIABLE = stringr::str_extract(name,".*(?=_\\d{3})"), # regex lookahead for '_001'
                     VARIABLE_SUBTOTAL = name,
                     VARIABLE_SUBTOTAL_DESC = label,
                     CONCEPT = concept) %>%
    dplyr::inner_join(acs_tables, by = "VARIABLE")


  # FIGURE OUT THE VARIABLE_ROLE INDICATORS

  population_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "POPULATION") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = "total"
    )


  race_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "RACE") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Total$") ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Not Hispanic or Latino!!White alone") ~ "omit",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Latino$") ~ "omit",
                       TRUE ~ "count"

                     )
    )

  ed_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "EDUCATION") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Total$") ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "No schooling") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "grade") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "High school") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Some college") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Associate's degree") ~ "count",
                       TRUE ~ "omit"

                     ))

  inc_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "INCOME") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Total$") ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Less") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$14,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$19,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$24,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$29,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$34,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$39,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$44,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$49,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$59,999") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "\\$74,999") ~ "count",
                       TRUE ~ "omit"

                     ))

  tenure_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "TENURE") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Total$") ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Renter occupied$") ~ "count",
                       TRUE ~ "omit"

                     ))


  burden_own_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "OWNER_COST_BURDEN") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       VARIABLE_SUBTOTAL_DESC %in% "Estimate!!Total!!Owner-occupied housing units" ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Owner") & stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "30 percent") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Owner") & stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Zero or negative income") ~ "count",
                       TRUE ~ "omit"

                     ))

  burden_rent_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "RENTER_COST_BURDEN") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR,
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       VARIABLE_SUBTOTAL_DESC %in% "Estimate!!Total!!Renter-occupied housing units" ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Renter") & stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "30 percent") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Renter") & stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Zero or negative income") ~ "count",
                       TRUE ~ "omit"

                     ))

  rent_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "RENT") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR = "RENT",
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = "include"
    )

  value_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "VALUE") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR = "VALUE",
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = "include"
    )

  median_inc_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "INCOME" & MEASURE_TYPE %in% c("MEDIAN")) %>%
    dplyr::transmute(SOURCE,
                     INDICATOR = "INCOME",
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = "include"
    )

  multifamily_vars_join <- all_vars %>%
    dplyr::filter(INDICATOR %in% "MULTIFAMILY") %>%
    dplyr::transmute(SOURCE,
                     INDICATOR = "MULTIFAMILY",
                     VARIABLE_SUBTOTAL,
                     VARIABLE_ROLE = dplyr::case_when(
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "Total$") ~ "total",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "2$") ~ "count", # remember: this is **count** data not value data ($ is for the end of the string)
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "3 or 4$") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "5 to 9$") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "10 to 19$") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "20 to 49$") ~ "count",
                       stringr::str_detect(VARIABLE_SUBTOTAL_DESC, "50 or more$") ~ "count",
                       TRUE ~ "omit")
    )

  acs_vars_join <- list(population_vars_join,
                        race_vars_join,
                        ed_vars_join,
                        inc_vars_join,
                        tenure_vars_join,
                        burden_own_vars_join,
                        burden_rent_vars_join,
                        rent_vars_join,
                        value_vars_join,
                        median_inc_vars_join,
                        multifamily_vars_join) %>%
    purrr::reduce(dplyr::bind_rows)


  # JOIN ROLES TO ACS DATA --------------------------------------------------

  acs_vars_data <- acs_data %>%
    dplyr::full_join(acs_vars_join, by = c("SOURCE", "VARIABLE_SUBTOTAL"))


  # ADJUST FOR INFLATION ----------------------------------------------------

  acs_vars_data_2018_dollars <- acs_vars_data %>%
    dplyr::mutate(ESTIMATE = dplyr::case_when(
      # only adjust ESTIMATE for the median income INDICATOR
     INDICATOR %in% c("INCOME") & MEASURE_TYPE %in% c("MEDIAN") ~ purrr::map2_dbl(ESTIMATE,
                                                            DATE_END,
                                                            convert_to_2018_dollars,
                                                            cpi = cpi,
                                                            series_title = "all"),
     # only adjust ESTIMATE for the price-related INDICATORS
     INDICATOR %in% c("RENT", "VALUE") ~ purrr::map2_dbl(ESTIMATE,
                                                            DATE_END,
                                                            convert_to_2018_dollars,
                                                            cpi = cpi,
                                                            series_title = "less_shelter"),
     TRUE ~ ESTIMATE)
    )



  # REFINE VARIABLE AND CREATE VARIABLE_DESC --------------------------------

  acs_vars_data <- acs_vars_data_2018_dollars %>%
    dplyr::mutate(VARIABLE = dplyr::case_when( # the cost burden vars need to be distinguishable from one another
      INDICATOR %in% "OWNER_COST_BURDEN" ~ stringr::str_c(VARIABLE,"_OWN"),
      INDICATOR %in% "RENTER_COST_BURDEN" ~ stringr::str_c(VARIABLE,"_RENT"),
      TRUE ~ VARIABLE
    )) %>%
    dplyr::mutate(VARIABLE_DESC = stringr::str_c(INDICATOR, SOURCE, sep = "_"))

  # ARRANGE COLUMNS WITH TEMPLATE -------------------------------------------

  acs_vars_ready <- variable_template %>%
    dplyr::full_join(acs_vars_data,
                     by = c("SOURCE",
                            "GEOGRAPHY_ID",
                            "GEOGRAPHY_ID_TYPE",
                            "GEOGRAPHY_NAME",
                            "GEOGRAPHY_TYPE",
                            "DATE_GROUP_ID",
                            "DATE_BEGIN",
                            "DATE_END",
                            "DATE_RANGE",
                            "DATE_RANGE_TYPE",
                            "INDICATOR",
                            "VARIABLE",
                            "VARIABLE_DESC",
                            "VARIABLE_SUBTOTAL",
                            "VARIABLE_SUBTOTAL_DESC",
                            "VARIABLE_ROLE",
                            "MEASURE_TYPE",
                            "ESTIMATE",
                            "MOE"))

  acs_variables <- acs_vars_ready

  # CHECK DATA --------------------------------------------------------------


  check_acs_vars_ready <- function(){

    # This function shows all of the INDICATOR values and their INDICATOR_ROLEs.
    # If any NA's are showing up then something needs to be fixed

    acs_variables %>% dplyr::count(DATE_GROUP_ID, INDICATOR, VARIABLE, VARIABLE_ROLE) %>% print(n=Inf)
  }



  # RETURN ------------------------------------------------------------------

  return(acs_variables)

}