refactor/add mock data functions

Author: andrewallenbruce

DESCRIPTION

@@ -17,21 +17,21 @@ Imports:
     collapse,
     dplyr,
     fixtuRes,
+    forcats,
     fs,
     lubridate,
     pins,
     purrr,
     rlang,
     santoku,
+    stringr,
+    tidyr,
     vctrs,
     wakefield,
-    withr (>= 3.0.0),
-    ymd
+    withr (>= 3.0.0)
 Suggests:
     covr,
-    forcats,
     fuimus,
-    ggforce,
     ggplot2,
     ggrepel,
     ggthemes,
@@ -46,7 +46,7 @@ Suggests:
     scales,
     svglite,
     testthat (>= 3.0.0),
-    tidyr
+    ymd
 VignetteBuilder: 
     knitr
 Remotes: 

NAMESPACE

@@ -11,6 +11,8 @@ export(mock_claims)
 export(mock_parbx)
 export(net_ex)
 export(predict_net)
+export(prep_claims)
 export(remove_quiet)
 export(sorted_bars)
 export(sum_na)
+export(summarise_claims)

R/generated-globals.R

@@ -9,23 +9,45 @@ utils::globalVariables(c(
   "age",
   # <bin_aging>
   # <mock_parbx>
+  # <prep_claims>
   "aging_bin",
   # <mock_claims>
+  # <prep_claims>
+  # <summarise_claims>
   "balance",
   # <mock_claims>
+  # <prep_claims>
+  # <summarise_claims>
   "charges",
+  # <prep_claims>
+  "claimid",
+  # <prep_claims>
+  # <summarise_claims>
+  "dar",
   # <mock_claims>
+  # <prep_claims>
   "date_acceptance",
   # <mock_claims>
+  # <prep_claims>
   "date_adjudication",
   # <mock_claims>
+  # <prep_claims>
   "date_reconciliation",
   # <mock_claims>
+  # <prep_claims>
   "date_release",
   # <mock_claims>
+  # <prep_claims>
   "date_service",
   # <mock_claims>
+  # <prep_claims>
   "date_submission",
+  # <prep_claims>
+  "date_type",
+  # <prep_claims>
+  "dates",
+  # <prep_claims>
+  "days",
   # <predict_net>
   "earb_gt120",
   # <predict_net>
@@ -48,6 +70,8 @@ utils::globalVariables(c(
   "net_pred",
   # <predict_net>
   "parl_120",
+  # <prep_claims>
+  "payer",
   # <predict_net>
   "pct_paid",
   # <avg_dar>

R/mock_data.R

@@ -23,7 +23,7 @@ mock_claims <- function(rows = 100, payers = payer_names(), count_days = FALSE,
   rsmpl <- sample(1:rows, size = (75 * rows / 100))
 
   df <- dplyr::tibble(
-    claimid             = wakefield::id(n = rows),
+    claimid             = as.character(wakefield::id(n = rows)),
     payer               = forcats::as_factor(fixtuRes::set_vector(size = rows, set = payers)),
     charges             = as.double(wakefield::income(n = rows, digits = 2) / 300),
     age                 = as.double(wakefield::age(n = rows, x = 15:100)),
@@ -131,29 +131,29 @@ payer_names <- function() {
     "Medicare",
     "Medicaid",
     "Kaiser Permanente",
-    "Elevance Health (Anthem)",
-    "HCSC (Health Care Service Corp)",
+    "Elevance (Anthem)",
+    "HCSC",
     "UnitedHealth",
     "Centene",
-    "CVS Health (Aetna)",
+    "CVS Aetna",
     "Humana",
-    "Cigna Health",
-    "Molina Healthcare",
-    "GuideWell (Florida Blue)",
+    "Cigna",
+    "Molina",
+    "GuideWell",
     "Highmark",
-    "BCBS Michigan",
-    "University Health Care",
-    "BCBS Wyoming",
-    "Bright Healthcare of Texas",
-    "Oscar Insurance Company",
-    "Wellcare, Inc.",
-    "Omaha Supplemental",
-    "Athene Annuity and Life",
-    "American General",
-    "Massachusetts Mutual Life",
+    "BCBS MI",
+    "Univ. Healthcare",
+    "BCBS WY",
+    "Bright",
+    "Oscar",
+    "Wellcare",
+    "Omaha",
+    "Athene",
+    "American",
+    "Mass Mutual",
     "New York Life",
-    "Lincoln National",
-    "Equitable Financial",
-    "Allianz Insurance"
+    "Lincoln Nat'l",
+    "Equitable",
+    "Allianz"
   )
 }

R/summarise_claims.R

@@ -0,0 +1,110 @@
+#' Prep mock coding/billing data frame
+#'
+#' @param df `[data.frame]` data frame generated by `mock_claims()`
+#'
+#' @param ... `[dots]` additional arguments
+#'
+#' @returns A [tibble][tibble::tibble-package]
+#'
+#' @examples
+#' prep_claims(mock_claims(rows = 5))
+#'
+#' @autoglobal
+#'
+#' @export
+prep_claims <- function(df, ...) {
+
+  df |>
+    tidyr::pivot_longer(
+      cols         = dplyr::starts_with("date"),
+      names_to     = "date_type",
+      names_prefix = "date_",
+      values_to    = "date") |>
+    dplyr::mutate(
+      date_type = forcats::fct_relevel(
+        date_type,
+        "service",
+        "release",
+        "submission",
+        "acceptance",
+        "adjudication",
+        "reconciliation"),
+      days = as.numeric(dplyr::lead(date) - date),
+      days = dplyr::lag(days, order_by = date),
+      .by = claimid) |>
+    dplyr::arrange(claimid, date_type) |>
+    tidyr::pivot_wider(
+      names_from = date_type,
+      names_glue = "{.value}_{date_type}",
+      values_from = c(date, days)
+    ) |>
+    remove_quiet() |>
+    dplyr::rowwise() |>
+    dplyr::mutate(dar = dplyr::if_else(
+      !is.na(date_reconciliation),
+      as.numeric(date_reconciliation - date_service),
+      as.numeric(date_adjudication - date_service)
+    )) |>
+    dplyr::ungroup() |>
+    tidyr::nest(
+      dates = c(
+        date_release,
+        date_submission,
+        date_acceptance,
+        date_adjudication,
+        date_reconciliation)
+    ) |>
+    bin_aging(dar) |>
+    dplyr::select(
+      claimid,
+      payer,
+      charges,
+      balance,
+      date_service,
+      aging_bin,
+      dar,
+      dplyr::starts_with("days_"),
+      dates
+    ) |>
+    .add_class()
+}
+
+#' Summarise mock coding/billing data frame
+#'
+#' @param df `[data.frame]` data frame generated from `mock_claims() |> prep_claims()`
+#'
+#' @param vars `[character]` variables to summarise, e.g. `c(days_service, days_reconciliation)`
+#'
+#' @returns A [tibble][tibble::tibble-package]
+#'
+#' @examples
+#' x <- mock_claims(rows = 500) |> prep_claims()
+#'
+#' summarise_claims(x)
+#'
+#' x |>
+#'   dplyr::group_by(
+#'     year = ymd::year(date_service),
+#'     month = ymd::month(date_service)) |>
+#'   summarise_claims()
+#'
+#' @autoglobal
+#'
+#' @export
+summarise_claims <- function(df, vars = c(dplyr::starts_with("days_"), dar)) {
+
+  df |>
+    dplyr::summarise(
+      n_claims      = dplyr::n(),
+      gross_charges = sum_na(charges),
+      ending_ar     = sum_na(balance),
+      dplyr::across({{ vars }}, \(x) mean_na(x), .names = "mean_{.col}"),
+      .groups = "drop"
+    ) |>
+    dplyr::rename_with(
+      ~ stringr::str_replace(.x, "days_", ""),
+      dplyr::starts_with("mean_")
+    ) |>
+    .add_class()
+
+}

README.Rmd

@@ -57,7 +57,7 @@ pak::pak("andrewallenbruce/forager")
 
 ```{r setup, echo=TRUE, message=FALSE, warning=FALSE}
 library(tidyverse)
-library(clock)
+library(ymd)
 library(forager)
 library(fuimus)
 ```
@@ -71,78 +71,28 @@ library(fuimus)
 <br>
 
 ```{r echo=TRUE}
-long <- x |> 
-  pivot_longer(
-    cols      = starts_with("date"), 
-    names_to  = "date_type", 
-    names_prefix = "date_",
-    values_to = "date") |> 
-  mutate(
-    date_type = fct_relevel(
-      date_type, 
-      "service", 
-      "release",
-      "submission", 
-      "acceptance", 
-      "adjudication", 
-      "reconciliation"), 
-    days = as.numeric(lead(date) - date),
-    days = lag(days, order_by = date),
-    .by = claimid) |> 
-  arrange(claimid, date_type)
-
-long
+(x <- prep_claims(x))
 ```
 
 <br>
 
 ```{r}
-x <- long |> 
-  pivot_wider(
-    names_from = date_type,
-    names_glue = "{.value}_{date_type}",
-    values_from = c(date, days)
-  ) |> 
-  remove_quiet() |> 
-  rowwise() |> 
-  mutate(days_in_ar = if_else(
-      !is.na(date_reconciliation), 
-      as.numeric((date_reconciliation - date_service)),
-      as.numeric((date_adjudication - date_service))
-  )) |> 
-  ungroup() |> 
-  nest(dates = c(
-    date_release, 
-    date_submission, 
-    date_acceptance, 
-    date_adjudication, 
-    date_reconciliation
-    )
-  ) |>
-  bin_aging(days_in_ar)
-x
+summarise_claims(x) |> 
+  glimpse()
 ```
 
 
+
 ```{r echo=TRUE}
 x |> 
   group_by(
     year = year(date_service),
     month = month(date_service),
     payer
   ) |>
-  summarise(
-    claims             = n(), 
-    balance            = sum_na(balance),
-    avg_release        = mean_na(days_release), 
-    avg_submission     = mean_na(days_submission),
-    avg_acceptance     = mean_na(days_acceptance),
-    avg_adjudication   = mean_na(days_adjudication),
-    avg_reconciliation = mean_na(days_reconciliation),
-    avg_days_in_ar     = mean_na(days_in_ar),
-    .groups = "drop") |> 
+  summarise_claims() |> 
   arrange(payer) |>
-  select(year, month, payer, claims, balance, avg_days_in_ar)
+  select(year, month, payer, n_claims, ending_ar, mean_dar)
 ```
 
 <br>
@@ -154,18 +104,9 @@ x |>
     qtr = quarter(date_service),
     payer
     ) |>
-  summarise(
-    claims             = n(), 
-    balance            = sum_na(balance),
-    avg_release        = mean_na(days_release), 
-    avg_submission     = mean_na(days_submission),
-    avg_acceptance     = mean_na(days_acceptance),
-    avg_adjudication   = mean_na(days_adjudication),
-    avg_reconciliation = mean_na(days_reconciliation),
-    avg_days_in_ar     = mean_na(days_in_ar),
-    .groups = "drop") |> 
+  summarise_claims() |> 
   arrange(payer) |>
-  select(year, qtr, payer, claims, balance, avg_days_in_ar)
+  select(year, qtr, payer, n_claims, ending_ar, mean_dar)
 ```
 
 ## Days in AR Calculation