Closes #1697 add keep_source_vars arg@devel (#2070)

* #1697 Added `keep_vars_source` arg in `derive_extreme_records()`, added relevent unit tests and updated NEWS.md. * Updated pkgdown.yml * Undoing renv.lock update * #1697 Updated default from `NULL` to `exprs(everything())` * #1697 Updated wordlist to include 'tidyselect' * #1697 Moved `select()` after `process_set_values_to()` * #1697 Updated `keep_vars_source` to `keep_source_vars` in `derive_extreme_records()` * Update NEWS.md Co-authored-by: Zelos Zhu <[email protected]> * Update NEWS.md Co-authored-by: Zelos Zhu <[email protected]> * Update R/derive_extreme_records.R Co-authored-by: Zelos Zhu <[email protected]> * #1697 Reverted changes made to `derive_var_merged_exist_flag()` and updated `derive_extreme_records()` * #1697 Updated 'Details' section to include `keep_source_vars` --------- Co-authored-by: Zelos Zhu <[email protected]>
pharmaverse · Sep 7, 2023 · bdcfe12 · bdcfe12
1 parent c16d77d
commit bdcfe12
Show file tree

Hide file tree

Showing 5 changed files with 117 additions and 8 deletions.
diff --git a/NEWS.md b/NEWS.md
@@ -24,6 +24,8 @@ assessment. (#1960)
   created even if values contributing to the computed values are `NA`.
 
 - The function `derive_vars_dy()` is updated to avoid potential error when the input `dataset` with columns ending with `temp`. (#2012)
+- Argument `keep_source_vars` was added to `derive_extreme_records()` which
+specifies which variables in the new observations should be kept. (#1697) 
 
 - Templates, vignettes, and other uses of `{admiral.test}` SDTM data are updated to use `{pharmaversesdtm}` instead. (#2040)
 
@@ -189,6 +191,7 @@ now. This affects `derive_param_tte()`. (#1727)
     - The `analysis_value` argument was enhanced such that any variable of the
       form `<variable>.<parameter>` can be used, e.g., `QSORRES.CHSF13`.
 
+
 ## Breaking Changes
 
 - `create_query_data()` and `derive_vars_query()` updated to rename variables in 

diff --git a/R/derive_extreme_records.R b/R/derive_extreme_records.R
@@ -95,6 +95,14 @@
 #'
 #'   *Permitted Values*: a condition
 #'
+#' @param keep_source_vars Variables to be kept in the new records
+#'
+#'   A named list or tidyselect expressions created by `exprs()` defining the
+#'   variables to be kept for the new records. The variables specified for
+#'   `by_vars` and `set_values_to` need not be specified here as they are kept
+#'   automatically.
+#'
+#'
 #' @inheritParams filter_extreme
 #' @inheritParams derive_summary_records
 #'
@@ -109,6 +117,9 @@
 #'   but not in the selected records are added.
 #'   1. The variables specified by the `set_values_to` argument are added to
 #'   the selected observations.
+#'   1. The variables specified by the `keep_source_vars` argument are selected
+#'   along with the variables specified in `by_vars` and `set_values_to`
+#'   arguments.
 #'   1. The observations are added to input dataset.
 #'
 #'
@@ -138,12 +149,14 @@
 #' # Add a new record for each USUBJID storing the minimum value (first AVAL).
 #' # If multiple records meet the minimum criterion, take the first value by
 #' # AVISITN. Set AVISITN = 97 and DTYPE = MINIMUM for these new records.
+#' # Specify the variables that need to be kept in the new records.
 #' derive_extreme_records(
 #'   adlb,
 #'   by_vars = exprs(USUBJID),
 #'   order = exprs(AVAL, AVISITN),
 #'   mode = "first",
 #'   filter_add = !is.na(AVAL),
+#'   keep_source_vars = exprs(AVAL),
 #'   set_values_to = exprs(
 #'     AVISITN = 97,
 #'     DTYPE = "MINIMUM"
@@ -254,6 +267,7 @@ derive_extreme_records <- function(dataset = NULL,
                                    exist_flag = NULL,
                                    true_value = "Y",
                                    false_value = "N",
+                                   keep_source_vars = exprs(everything()),
                                    set_values_to,
                                    filter) {
   if (!missing(filter)) {
@@ -268,6 +282,8 @@ derive_extreme_records <- function(dataset = NULL,
   # Check input arguments
   assert_vars(by_vars, optional = is.null(dataset_ref))
   assert_expr_list(order, optional = TRUE)
+  assert_expr_list(keep_source_vars, optional = TRUE)
+
   assert_data_frame(
     dataset,
     required_vars = expr_c(
@@ -333,10 +349,12 @@ derive_extreme_records <- function(dataset = NULL,
     new_obs <- new_add_obs
   }
 
-  new_obs <- process_set_values_to(
-    new_obs,
-    set_values_to = set_values_to
-  )
+  new_obs <- new_obs %>%
+    process_set_values_to(
+      set_values_to = set_values_to
+    ) %>%
+    select(!!!by_vars, names(set_values_to), !!!keep_source_vars)
+
 
   # Create output dataset
   bind_rows(dataset, new_obs)

diff --git a/inst/WORDLIST b/inst/WORDLIST
@@ -33,7 +33,6 @@ ATOXGR
 AVAL
 AVISITN
 Abercrombie
-
 Akinyi
 Alkalosis
 Alanine
@@ -184,7 +183,6 @@ Pooja
 Pre
 QD
 QID
-Quosures
 README
 Rahal
 Ratford
@@ -230,6 +228,7 @@ Thanikachalam
 Timepoint
 ULN
 USUBJID
+USUBJIDs
 Upadhyay
 VAD
 Vignesh
@@ -307,7 +306,6 @@ mmHg
 mmol
 modularized
 msec
-nolint
 occds
 onwards
 parttime
@@ -325,14 +323,14 @@ rlang
 roche
 roxygen
 scalable
-str
 submittable
 subperiod
 subperiods
 subscale
 summarization
 th
 thromboplastin
+tidyselect
 tidyverse
 tidyselect
 timeframe

diff --git a/man/derive_extreme_records.Rd b/man/derive_extreme_records.Rd
diff --git a/tests/testthat/test-derive_extreme_records.R b/tests/testthat/test-derive_extreme_records.R
@@ -376,3 +376,80 @@ test_that("derive_extreme_records Test 7: error if no input data", {
     fixed = TRUE
   )
 })
+
+## Test 8: keep vars in `keep_source_vars` in the new records ----
+test_that("derive_extreme_records Test 8: keep vars in `keep_source_vars` in the new records", {
+  input <- tibble::tribble(
+    ~USUBJID, ~AVISITN, ~AVAL, ~LBSEQ,
+    1, 1, 12, 1,
+    1, 3, 9, 2,
+    2, 2, 42, 1,
+    3, 3, 14, 1,
+    3, 3, 10, 2
+  )
+
+  expected_output <- bind_rows(
+    input,
+    tibble::tribble(
+      ~USUBJID, ~AVISITN, ~AVAL, ~LBSEQ,
+      1, 3, 9, 2,
+      2, 2, 42, 1,
+      3, 3, 10, 2
+    ) %>%
+      select(USUBJID, AVISITN, AVAL) %>%
+      mutate(DTYPE = "LOV")
+  )
+
+  actual_output <- derive_extreme_records(
+    input,
+    order = exprs(AVISITN, LBSEQ),
+    by_vars = exprs(USUBJID),
+    mode = "last",
+    keep_source_vars = exprs(AVISITN, AVAL),
+    set_values_to = exprs(DTYPE = "LOV")
+  )
+
+  expect_dfs_equal(
+    base = expected_output,
+    compare = actual_output,
+    keys = c("USUBJID", "AVISITN", "LBSEQ", "DTYPE")
+  )
+})
+
+## Test 9: keep all vars in the new records when `keep_source_vars` is 'exprs(everything())' ----
+test_that("derive_extreme_records Test 9: keep all vars in the new records when `keep_source_vars` is 'exprs(everything())'", { # nolint
+  input <- tibble::tribble(
+    ~USUBJID, ~AVISITN, ~AVAL, ~LBSEQ,
+    1, 1, 12, 1,
+    1, 3, 9, 2,
+    2, 2, 42, 1,
+    3, 3, 14, 1,
+    3, 3, 10, 2
+  )
+
+  expected_output <- bind_rows(
+    input,
+    tibble::tribble(
+      ~USUBJID, ~AVISITN, ~AVAL, ~LBSEQ,
+      1, 3, 9, 2,
+      2, 2, 42, 1,
+      3, 3, 10, 2
+    ) %>%
+      mutate(DTYPE = "LOV")
+  )
+
+  actual_output <- derive_extreme_records(
+    input,
+    order = exprs(AVISITN, LBSEQ),
+    by_vars = exprs(USUBJID),
+    mode = "last",
+    keep_source_vars = exprs(everything()),
+    set_values_to = exprs(DTYPE = "LOV")
+  )
+
+  expect_dfs_equal(
+    base = expected_output,
+    compare = actual_output,
+    keys = c("USUBJID", "AVISITN", "LBSEQ", "DTYPE")
+  )
+})