From d09efcd7e736d14937ee7bdf1ab13b4c3caadbd6 Mon Sep 17 00:00:00 2001
From: James McMahon <james.mcmahon@phs.scot>
Date: Tue, 18 Jul 2023 08:15:11 +0100
Subject: [PATCH] Make episode file output with `anon_chi` (#747)

I've added this as a parameter so you can output CHI if desired, but the default is for anon_chi.

For the tests, it swaps back to CHI as there are some tests which specifically us the CHI number.

Co-authored-by: Jennit07 <67372904+Jennit07@users.noreply.github.com>
---
 R/process_tests_episode_file.R |  5 +++--
 R/run_episode_file.R           | 16 +++++++++++++++-
 man/run_episode_file.Rd        | 10 +++++++++-
 3 files changed, 27 insertions(+), 4 deletions(-)

diff --git a/R/process_tests_episode_file.R b/R/process_tests_episode_file.R
index b595d1d54..46e9e7171 100644
--- a/R/process_tests_episode_file.R
+++ b/R/process_tests_episode_file.R
@@ -10,7 +10,7 @@ process_tests_episode_file <- function(data, year) {
   data <- data %>%
     dplyr::select(
       "year",
-      "chi",
+      "anon_chi",
       "gender",
       "postcode",
       "hbtreatcode",
@@ -20,7 +20,8 @@ process_tests_episode_file <- function(data, year) {
       "record_keydate1",
       "record_keydate2",
       dplyr::contains(c("beddays", "cost", "cij"))
-    )
+    ) %>%
+    slfhelper::get_chi()
 
   old_data <- get_existing_data_for_tests(data)
 
diff --git a/R/run_episode_file.R b/R/run_episode_file.R
index 45a4e6ed5..1f2bb33ed 100644
--- a/R/run_episode_file.R
+++ b/R/run_episode_file.R
@@ -4,11 +4,17 @@
 #' @param year The year to process, in FY format.
 #' @param write_to_disk (optional) Should the data be written to disk default is
 #' `TRUE` i.e. write the data to disk.
+#' @param anon_chi_out (Default:TRUE) Should `anon_chi` be used in the output
+#' (instead of chi)
 #'
 #' @return a [tibble][tibble::tibble-package] containing the episode file
 #' @export
 #'
-run_episode_file <- function(processed_data_list, year, write_to_disk = TRUE) {
+run_episode_file <- function(
+    processed_data_list,
+    year,
+    write_to_disk = TRUE,
+    anon_chi_out = TRUE) {
   episode_file <- dplyr::bind_rows(processed_data_list) %>%
     create_cost_inc_dna() %>%
     apply_cost_uplift() %>%
@@ -103,6 +109,14 @@ run_episode_file <- function(processed_data_list, year, write_to_disk = TRUE) {
     join_deaths_data(year) %>%
     load_ep_file_vars(year)
 
+  if (anon_chi_out) {
+    episode_file <- slfhelper::get_anon_chi(
+      episode_file,
+      chi_var = "chi",
+      drop = TRUE
+    )
+  }
+
   if (write_to_disk) {
     slf_path <- get_file_path(
       get_year_dir(year),
diff --git a/man/run_episode_file.Rd b/man/run_episode_file.Rd
index e85621b59..59d5fea1d 100644
--- a/man/run_episode_file.Rd
+++ b/man/run_episode_file.Rd
@@ -4,7 +4,12 @@
 \alias{run_episode_file}
 \title{Produce the Source Episode file}
 \usage{
-run_episode_file(processed_data_list, year, write_to_disk = TRUE)
+run_episode_file(
+  processed_data_list,
+  year,
+  write_to_disk = TRUE,
+  anon_chi_out = TRUE
+)
 }
 \arguments{
 \item{processed_data_list}{containing data from processed extracts.}
@@ -13,6 +18,9 @@ run_episode_file(processed_data_list, year, write_to_disk = TRUE)
 
 \item{write_to_disk}{(optional) Should the data be written to disk default is
 \code{TRUE} i.e. write the data to disk.}
+
+\item{anon_chi_out}{(Default:TRUE) Should \code{anon_chi} be used in the output
+(instead of chi)}
 }
 \value{
 a \link[tibble:tibble-package]{tibble} containing the episode file