refactor: aging_bins, avg_dar, generate_data

DESCRIPTION

@@ -27,6 +27,7 @@ Suggests:
     covr,
     tidyr,
     svglite,
+    janitor,
     ggplot2,
     headliner,
     forcats,

NAMESPACE

@@ -3,6 +3,7 @@
 export(avg_dar)
 export(bin_aging)
 export(dar_ex)
+export(days_between)
 export(generate_data)
 export(load_ex)
 export(net_ex)

R/aging.R

@@ -2,89 +2,115 @@
 #'
 #' @template args-df-default
 #'
-#' @template args-date-col
+#' @param ndays `<dbl>` column of counts of days elapsed to bin by
 #'
 #' @param bin_type `<chr>` string specifying the bin type; one of "chop", "cut" or "ivs"
 #'
 #' @template returns-default
 #'
 #' @examples
-#' binned <- bin_aging(
-#'   df = load_ex("aging_ex"),
-#'   date = dos
-#' ) |>
-#'   dplyr::select(
-#'     dos:ins_class,
-#'     dar:aging_bin
-#'   )
+#' generate_data(10)[c(
+#'   "date_srvc",
+#'   "charges",
+#'   "payer")] |>
+#'   days_between(date_srvc) |>
+#'   bin_aging(days_in_ar)
 #'
-#' head(binned)
-#'
-#' binned |>
+#' load_ex("aging_ex") |>
+#'   dplyr::select(dos, charges, ins_name) |>
+#'   days_between(dos) |>
+#'   bin_aging(days_in_ar) |>
 #'   dplyr::arrange(aging_bin) |>
-#'   dplyr::summarise(n_claims = dplyr::n(),
-#'                    balance = sum(charges),
-#'                    .by = aging_bin) |>
-#'   dplyr::mutate(
-#'     tot_claims = sum(n_claims),
-#'     tot_balance = sum(balance),
-#'     pct_claims = n_claims / tot_claims,
-#'     pct_balance = balance / tot_balance) |>
-#'   print(n = 50)
-#'
-#' binned |>
-#'   dplyr::arrange(aging_bin, ins_name) |>
-#'   dplyr::summarise(
-#'     n_claims = dplyr::n(),
-#'     balance = sum(charges),
-#'     .by = c(aging_bin, ins_name)
-#'   ) |>
-#'   dplyr::mutate(
-#'     tot_claims = sum(n_claims),
-#'     tot_balance = sum(balance),
-#'     pct_claims = n_claims / tot_claims,
-#'     pct_balance = balance / tot_balance) |>
-#'   print(n = 50)
-#'
-#' binned |>
-#'   dplyr::arrange(ins_name, aging_bin) |>
-#'   dplyr::summarise(
-#'     n_claims = dplyr::n(),
-#'     balance = sum(charges),
-#'     .by = c(aging_bin, ins_name)
+#'   dplyr::group_by(
+#'     year = clock::get_year(dos),
+#'     month = clock::date_month_factor(dos),
 #'   ) |>
-#'   dplyr::mutate(
-#'     tot_claims = sum(n_claims),
-#'     tot_balance = sum(balance),
-#'     pct_claims = n_claims / tot_claims,
-#'     pct_balance = balance / tot_balance,
-#'     .by = ins_name) |>
-#'   print(n = 50)
+#'   janitor::tabyl(ins_name, aging_bin, year)
 #'
 #' @autoglobal
 #'
 #' @export
-bin_aging <- function(df, date, bin_type = "chop") {
+bin_aging <- function(df, ndays, bin_type = c("case", "chop")) {
 
-  df <- df |>
-    dplyr::mutate(
-      dar = clock::date_count_between(
-        {{ date }},
-        lubridate::today(),
-        "day"))
+  bin_type <- match.arg(bin_type)
 
   if (bin_type == "chop") {
 
-  df <- df |>
-    dplyr::mutate(
-      aging_bin = santoku::chop_width(
-        dar,
-        30,
-        start = 0,
-        left = FALSE,
-        close_end = FALSE
+    df <- df |>
+      dplyr::mutate(
+        aging_bin = santoku::chop_width(
+          {{ ndays }},
+          30,
+          start = 0,
+          left = FALSE,
+          close_end = FALSE
+        )
+      )
+  }
+
+  if (bin_type == "case") {
+
+    df <- df |>
+      dplyr::mutate(
+        aging_bin = dplyr::case_when(
+          dplyr::between({{ ndays }}, 0, 30) ~ "0-30",
+          dplyr::between({{ ndays }}, 31, 60) ~ "31-60",
+          dplyr::between({{ ndays }}, 61, 90) ~ "61-90",
+          dplyr::between({{ ndays }}, 91, 120) ~ "91-120",
+          {{ ndays }} >= 121 ~ "121+"
+        ),
+        aging_bin = suppressWarnings(
+          forcats::fct_relevel(
+            aging_bin,
+            c("0-30", "31-60", "61-90", "91-120", "121+"),
+            after = Inf
+          )
+        )
       )
-    )
   }
   return(df)
 }
+
+#' Calculate Number of Days Between Two Dates
+#'
+#' @template args-df-default
+#'
+#' @param from `[character]` column of dates to calculate days between
+#'
+#' @param to `[character]` column of dates to calculate days between
+#'
+#' @template returns-default
+#'
+#' @examples
+#' generate_data(10)[c(
+#'   "date_srvc",
+#'   "charges",
+#'   "payer")] |>
+#'   days_between(date_srvc)
+#'
+#' @autoglobal
+#'
+#' @export
+days_between <- function(df, from, to = NULL) {
+
+  if (is.null(to)) {
+
+    df |>
+      dplyr::mutate(
+        days_in_ar = clock::date_count_between(
+          {{ from }},
+          clock::date_today(""),
+          "day")
+        )
+
+  } else {
+
+    df |>
+      dplyr::mutate(
+        days_in_ar = clock::date_count_between(
+          {{ from }},
+          {{ to }},
+          "day")
+        )
+    }
+}

R/dar.R

@@ -8,9 +8,9 @@
 #'
 #' @template args-earb-col
 #'
-#' @param dart `<dbl>` Target Days in AR, default is `35` days
+#' @param dart `[numeric]` Target Days in AR, default is `35` days
 #'
-#' @param period `<chr>` string specifying the calculation period; one of
+#' @param by `[character]` string specifying the calculation period; one of
 #'   `"month"`, `"quarter"`, or `"year"`; defaults to `"month"`
 #'
 #' @template returns-default
@@ -21,14 +21,14 @@
 #'         gct    = gross_charges,
 #'         earb   = ending_ar,
 #'         dart   = 35,
-#'         period = "month")
+#'         by = "month")
 #'
 #' avg_dar(df     = dar_ex(),
 #'         date   = date,
 #'         gct    = gross_charges,
 #'         earb   = ending_ar,
 #'         dart   = 35,
-#'         period = "quarter")
+#'         by = "quarter")
 #'
 #' @autoglobal
 #'
@@ -38,9 +38,9 @@ avg_dar <- function(df,
                     gct,
                     earb,
                     dart = 35,
-                    period = c("month", "quarter")) {
+                    by = c("month", "quarter")) {
 
-  period  <- match.arg(period)
+  by  <- match.arg(by)
   datecol <- rlang::englue("{{ date }}")
   earbcol <- rlang::englue("{{ earb }}")
   gctcol  <- rlang::englue("{{ gct }}")
@@ -63,7 +63,7 @@ avg_dar <- function(df,
     ndip         = lubridate::days_in_month({{ date }})
   )
 
-  if (period == "quarter") {
+  if (by == "quarter") {
 
     qtr_max_nmons <- df |>
       dplyr::summarise(

R/generated-globals.R

@@ -7,16 +7,25 @@ utils::globalVariables(c(
   # <avg_dar>
   "adc",
   # <bin_aging>
+  "aging_bin",
+  # <generate_data>
+  "balance",
+  # <generate_data>
+  "charges",
   # <avg_dar>
   "dar",
   # <generate_data>
-  "date_of_acceptance",
+  "date_accept",
+  # <generate_data>
+  "date_adjud",
   # <generate_data>
-  "date_of_release",
+  "date_recon",
   # <generate_data>
-  "date_of_service",
+  "date_rlse",
   # <generate_data>
-  "date_of_submission",
+  "date_srvc",
+  # <generate_data>
+  "date_submit",
   # <predict_net>
   "earb_gt120",
   # <predict_net>
@@ -29,6 +38,8 @@ utils::globalVariables(c(
   "max_nmon",
   # <avg_dar>
   "month",
+  # <generate_data>
+  "NA_Date_",
   # <avg_dar>
   "ndip",
   # <predict_net>

R/predict_net.R

@@ -30,16 +30,16 @@ predict_net <- function(df, date, gct, earb, net, parb_120) {
 
   df <- df |>
     dplyr::mutate(
-      date = lubridate::ymd({{ date }}),
-      pct_paid = {{ net }} / {{ gct }},
-      parl_120 = 1 - {{ parb_120 }},
-      net_pred = ({{ gct }} * pct_paid) * parl_120,
+      date       = lubridate::ymd({{ date }}),
+      pct_paid   = {{ net }} / {{ gct }},
+      parl_120   = 1 - {{ parb_120 }},
+      net_pred   = ({{ gct }} * pct_paid) * parl_120,
       earb_gt120 = {{ earb }} * parb_120,
       earb_lt120 = {{ earb }} * parl_120
     )
 
   pred <- dplyr::tibble(
-    date = df$date[nrow(df)] + months(1),
+    date     = df$date[nrow(df)] + months(1),
     net_pred = df$net_pred[nrow(df)]
   )
 
@@ -75,11 +75,12 @@ predict_net <- function(df, date, gct, earb, net, parb_120) {
 net_ex <- function() {
 
   load_ex("monthly_raw") |>
-    dplyr::select(date,
-                  gct = gross_charges,
-                  earb = ending_ar,
-                  net = net_payment,) |>
-    dplyr::mutate(date = lubridate::ymd(date),
-                  parb_120 = rep(c(0.021, 0.047, 0.075), 4)
-                  )
+    dplyr::select(
+      date,
+      gct      = gross_charges,
+      earb     = ending_ar,
+      net      = net_payment,) |>
+    dplyr::mutate(
+      date     = lubridate::ymd(date),
+      parb_120 = rep(c(0.021, 0.047, 0.075), 4))
 }