Allow numeric input to count_by in analyze_num_patients (#1141)

Fixes #1062

NEWS.md

@@ -19,6 +19,7 @@
 * Added `x_lab` parameter to `g_lineplot` to customize x-axis label.
 * Remove 25% padding of y-axis in `g_lineplot`.
 * Added support for creating multiple risk difference columns, each comparing to a single comparison group. Multiple comparison groups can be specified as a vector via the `arm_y` argument.
+* Allowed numeric vector as `count_by` input in `analyze_num_patients` and `summarize_num_patients`.
 
 ### Bug Fixes
 * Fixed bug in `decorate_grob` preventing text wrapping from accounting for font size.

R/summarize_num_patients.R

@@ -5,8 +5,8 @@
 #' Count the number of unique and non-unique patients in a column (variable).
 #'
 #' @inheritParams argument_convention
-#' @param count_by (`character` or `factor`)\cr optional vector to be combined with `x` when counting
-#'   `nonunique` records.
+#' @param count_by (`vector`)\cr optional vector of any type to be combined with `x` when counting `nonunique`
+#'   records.
 #' @param unique_count_suffix (`logical`)\cr should `"(n)"` suffix be added to `unique_count` labels.
 #'   Defaults to `TRUE`.
 #' @param .stats (`character`)\cr statistics to select for the table. Run `get_stats("summarize_num_patients")`
@@ -35,7 +35,7 @@ NULL
 #'   x = as.character(c(1, 1, 1, 2, 4, NA)),
 #'   labelstr = "",
 #'   .N_col = 6L,
-#'   count_by = as.character(c(1, 1, 2, 1, 1, 1))
+#'   count_by = c(1, 1, 2, 1, 1, 1)
 #' )
 #'
 #' @export
@@ -51,7 +51,6 @@ s_num_patients <- function(x, labelstr, .N_col, count_by = NULL, unique_count_su
 
   if (!is.null(count_by)) {
     checkmate::assert_vector(count_by, len = length(x))
-    checkmate::assert_multi_class(count_by, classes = c("factor", "character"))
     count2 <- n_available(unique(interaction(x, count_by)))
   }
 
@@ -86,7 +85,7 @@ s_num_patients <- function(x, labelstr, .N_col, count_by = NULL, unique_count_su
 #'
 #' df_by_event <- data.frame(
 #'   USUBJID = as.character(c(1, 2, 1, 4, NA)),
-#'   EVENT = as.character(c(10, 15, 10, 17, 8))
+#'   EVENT = c(10, 15, 10, 17, 8)
 #' )
 #' s_num_patients_content(df_by_event, .N_col = 5, .var = "USUBJID", count_by = "EVENT")
 #'
@@ -112,10 +111,7 @@ s_num_patients_content <- function(df,
   }
 
   x <- df[[.var]]
-  y <- switch(as.numeric(!is.null(count_by)) + 1,
-    NULL,
-    df[[count_by]]
-  )
+  y <- if (is.null(count_by)) NULL else df[[count_by]]
 
   s_num_patients(
     x = x,

tests/testthat/test-summarize_num_patients.R

@@ -83,7 +83,7 @@ testthat::test_that("summarize_num_patients works as expected with healthy input
 
 testthat::test_that("s_num_patients count_by works as expected with healthy input", {
   x <- as.character(c(1, 2, 1, 4, 1))
-  y <- as.character(c(6, 7, 8, 9, 6))
+  y <- c(6, 7, 8, 9, 6)
   result <- s_num_patients(x = x, labelstr = "", .N_col = 5, count_by = y)
 
   res <- testthat::expect_silent(result)
@@ -92,7 +92,7 @@ testthat::test_that("s_num_patients count_by works as expected with healthy inpu
 
 testthat::test_that("s_num_patients count_by with missing works as expected with healthy input", {
   x <- as.character(c(1, 2, 1, 4, NA))
-  y <- as.character(c(6, 7, 8, 9, 6))
+  y <- c(6, 7, 8, 9, 6)
   result <- s_num_patients(x = x, labelstr = "", .N_col = 5, count_by = y)
 
   res <- testthat::expect_silent(result)
@@ -101,7 +101,7 @@ testthat::test_that("s_num_patients count_by with missing works as expected with
 
 testthat::test_that("s_num_patients count_by with missing case 2 works as expected with healthy input", {
   x <- as.character(c(1, 2, 1, 4, 1))
-  y <- as.character(c(6, 7, NA, 9, 6))
+  y <- c(6, 7, NA, 9, 6)
   result <- s_num_patients(x = x, labelstr = "", .N_col = 5, count_by = y)
 
   res <- testthat::expect_silent(result)
@@ -111,7 +111,7 @@ testthat::test_that("s_num_patients count_by with missing case 2 works as expect
 testthat::test_that("s_num_patients_content with count_by works as expected with healthy input", {
   df <- data.frame(
     USUBJID = as.character(c(1, 2, 1, 4, NA)),
-    AGE = as.character(c(10, 15, 10, 17, 8))
+    AGE = c(10, 15, 10, 17, 8)
   )
   result <- s_num_patients_content(df = df, .N_col = 5, .var = "USUBJID", count_by = "AGE")
 
@@ -122,7 +122,7 @@ testthat::test_that("s_num_patients_content with count_by works as expected with
 testthat::test_that("s_num_patients_content with count_by case 2 works as expected with healthy input", {
   df <- data.frame(
     USUBJID = as.character(c(1, 2, 1, 4, NA)),
-    AGE = as.character(c(10, 15, 11, 17, 8))
+    AGE = c(10, 15, 11, 17, 8)
   )
   result <- s_num_patients_content(df = df, .N_col = 5, .var = "USUBJID", count_by = "AGE")
 
@@ -133,7 +133,7 @@ testthat::test_that("s_num_patients_content with count_by case 2 works as expect
 testthat::test_that("s_num_patients_content with count_by trivial cases, identical to without count_by", {
   df <- data.frame(
     USUBJID = as.character(c(1, 2, 1, 4, 9)),
-    AGE = as.character(c(10, 15, 11, 17, 8))
+    AGE = c(10, 15, 11, 17, 8)
   )
   result <- s_num_patients_content(df = df, .N_col = 5, .var = "USUBJID", count_by = "USUBJID")
 
@@ -145,7 +145,7 @@ testthat::test_that("summarize_num_patients with count_by works as expected with
   df <- data.frame(
     USUBJID = as.character(c(1, 2, 1, 4, NA, 6, 6, 8, 9)),
     ARM = c("A", "A", "A", "A", "A", "B", "B", "B", "B"),
-    BY = as.character(c(10, 15, 10, 17, 8, 11, 11, 19, 17))
+    BY = c(10, 15, 10, 17, 8, 11, 11, 19, 17)
   )
 
   # Check with both output