R/census2010_to_isco08.R

In Guidowe/occupationcross: Package for making crosswalks among different Occupational codes

#' Cross Census 2010 (USA) to ISCO 08
#'
#' @param base Dataframe from any survey including a variable with Standard Occupational Codes (SOC) 2010
#' @param census character vector containing Census 2010 codes
#' @param summary If TRUE provides other dataframe counting how many cases where asigned for each Census 08 code to each ISCO88 code.
#' @param code_titles If TRUE adds classification titles besides from codes.
#'
#' @return The function returns the provided dataframe, adding a new variable with the crosswalk to ISCO 08 codes
#' @export
#' @examples
#'
#' base_crossed <- census2010_to_isco08(toy_base_ipums_cps_2018,census = OCC2010,summary = TRUE)
#'
census2010_to_isco08<- function(base,
                                census,
                                summary = FALSE,
                                code_titles = FALSE){

base <-  base %>%
    dplyr::mutate(Census = as.character({{census}}))


Codigos_error <- base %>%
    dplyr::select(Census) %>%
    dplyr::filter(!(Census %in% unique(crosstable_census2010_soc2010$Census))) %>%
    unique()

if(length(Codigos_error$Census)>=1){
  warning(paste0("The following codes from the provided database were not found in 'crosstable_census2010_soc2010_isco08' and it was not possible to crosswalk them:  ",
                 list(Codigos_error$Census)))

  base  <- base %>%
    dplyr::mutate(
      Census = dplyr::case_when(
        Census %in% Codigos_error$Census ~ "0",
        TRUE~ Census))
}

nested.data.isco.cross <- crosstable_census2010_soc2010_isco08 %>%
    dplyr::group_by(Census,SOC) %>%
    tidyr::nest()

base_join  <- base %>%
    dplyr::mutate(Census = as.character(Census)) %>%
    dplyr::left_join(nested.data.isco.cross,by = "Census")

set.seed(999971)

sample.isco <- function(df) {
  sample(df$ISCO,size = 1)
}

base_join_sample <- base_join %>%
    dplyr::mutate(ISCO.08 = purrr::map(.x = data,
                                       .f = sample.isco))  %>%
    dplyr::select(-data) %>%
    dplyr::mutate(ISCO.08 = as.numeric(ISCO.08))

if (code_titles==TRUE) {

    titles  <- crosstable_census2010_soc2010_isco08 %>%
      dplyr::mutate(ISCO.08 = as.numeric(ISCO)) %>%
      dplyr::select(ISCO.08,ISCO.title) %>%
      unique()

    base_join_sample <- base_join_sample %>%
      dplyr::left_join(titles)
  }



return(base_join_sample)

if (summary==TRUE) {

    summary_cross <- base_join_sample %>%
      dplyr::group_by(Census,ISCO.08) %>%
      dplyr::summarise(Cases = dplyr::n())

print(summary_cross)

      }


}