R/01 - Create Transactional Flat DF.R

In mjkarlsen/traumaR: This package was created to manipulate the Pennsylvania Trauma Outcome System (PTOS) data.

#' Create Flat Dataframe of all procedures, complications, and diagnosis by ID.
#'   Uses the trans_full_df as its source
#'
#' @param .data
#'
#' @return
#' @export
#'
create_flat_trans <- function(.full_trans) {

  forearm_surgery <- list('79.02', '79.12', '79.22', '79.32') #forearm surgery
  forearm_diag <- list(seq(from = 813.00, to = 813.99, by = 0.01)) #diagnosis for forearm fx

  .flat_trans <- .full_trans %>%
    filter.(data_source != 'injury') %>%
    summarize.(code_cd = paste(code_cd, collapse = " "),
               code_desc = paste(code_desc, collapse = ", "),
               by = c(id, data_source)
    ) %>%
    pivot_wider.(names_from = data_source,
                 values_from = c(code_cd, code_desc)) %>%
    drop_na.() %>%
    mutate.(
      # fltr_diagnosis = (str_detect(code_cd_diagnosis, #diagnosis for forearm fx
      #                                       pattern = !!forearm_diag)),
      #       fltr_procedure = (str_detect(code_cd_procedure,
      #                                       pattern = !!forearm_surgery)),#forearm surgery
            fltr_complication = str_detect(code_cd_complication,
                                           pattern = '32'),
            fltr_fasciotomy = str_detect(code_cd_procedure,
                                         pattern = '83.14'))

  assign(paste("trans_flat_df"), data.frame(.flat_trans), envir = .GlobalEnv)
}