trigger test from yurovska/tern/tree/821_ancova_weights

.lintr

@@ -1,5 +1,4 @@
 linters: linters_with_defaults(
   line_length_linter = line_length_linter(120),
-  cyclocomp_linter = NULL,
   object_usage_linter = NULL
   )

DESCRIPTION

@@ -24,8 +24,8 @@ Depends:
     formatters (>= 0.5.10.9001),
     R (>= 4.2),
     rlistings (>= 0.2.10),
-    rtables (>= 0.6.11),
-    tern (>= 0.9.7.9003)
+    rtables (>= 0.6.11.9017),
+    tern (>= 0.9.7.9018)
 Imports:
     pharmaverseadam (>= 1.1.0),
     random.cdisc.data (>= 0.3.16)
@@ -46,6 +46,7 @@ VignetteBuilder:
     knitr
 Remotes:
     insightsengineering/formatters@main,
+    insightsengineering/rtables@main,
     insightsengineering/tern@main,
     pharmaverse/pharmaverseadam@main
 biocViews:

tests/testthat/_snaps/table_aovt01.md

@@ -63,3 +63,47 @@
             95% CI                                 (-1.17, 2.76)   (-0.56, 3.39)
             p-value                                   0.4288          0.1591    
 
+# AOVT01 variant with proportional weights is produced correctly
+
+    Code
+      res
+    Output
+                                          ARM A        ARM B           ARM C    
+                                         (N=134)      (N=134)         (N=132)   
+      ——————————————————————————————————————————————————————————————————————————
+      BFIALL                                                                    
+        Adjusted mean                                                           
+          n                                134          134             132     
+          Adjusted Mean                   4.50         6.36            4.05     
+          Difference in Adjusted Means                 1.85            -0.46    
+            95% CI                                 (-0.14, 3.85)   (-2.45, 1.54)
+            p-value                                   0.0679          0.6539    
+      FATIGI                                                                    
+        Adjusted mean                                                           
+          n                                134          134             132     
+          Adjusted Mean                   5.42         4.83            4.56     
+          Difference in Adjusted Means                 -0.59           -0.86    
+            95% CI                                 (-2.58, 1.41)   (-2.87, 1.15)
+            p-value                                   0.5644          0.4026    
+      FKSI-FWB                                                                  
+        Adjusted mean                                                           
+          n                                134          134             132     
+          Adjusted Mean                   4.31         3.53            3.08     
+          Difference in Adjusted Means                 -0.79           -1.24    
+            95% CI                                 (-2.71, 1.14)   (-3.17, 0.69)
+            p-value                                   0.4221          0.2088    
+      FKSI-TSE                                                                  
+        Adjusted mean                                                           
+          n                                134          134             132     
+          Adjusted Mean                   4.70         3.84            4.45     
+          Difference in Adjusted Means                 -0.86           -0.25    
+            95% CI                                 (-2.80, 1.09)   (-2.20, 1.70)
+            p-value                                   0.3858          0.8007    
+      FKSIALL                                                                   
+        Adjusted mean                                                           
+          n                                134          134             132     
+          Adjusted Mean                   5.00         5.79            6.42     
+          Difference in Adjusted Means                 0.79            1.42     
+            95% CI                                 (-1.17, 2.76)   (-0.56, 3.39)
+            p-value                                   0.4288          0.1591    
+

tests/testthat/test-char-support.R

@@ -0,0 +1,120 @@
+adsl <- random.cdisc.data::cadsl
+adeg <- random.cdisc.data::cadeg
+
+################################################################################
+# Define script level parameters:
+################################################################################
+adsl_f <- adsl %>%
+  select(USUBJID, ACTARM)
+
+# Note:  We keep only post-baseline for analysis.
+adeg_f <- adeg %>%
+  filter(AVISIT %in% levels(adeg$AVISIT)[1:3]) %>%
+  filter(PARAM %in% c("Heart Rate", "QT Duration", "RR Duration")) %>%
+  var_relabel(
+    PARAM = "Assessment",
+    ANRIND = "Abnormality"
+  ) %>%
+  select(USUBJID, AVISIT, PARAM, PARAMCD, ANRIND, ACTARM) %>%
+  mutate(PARCAT = case_when(
+    PARAMCD == "QT" ~ "QT",
+    TRUE ~ "Other"
+  )) %>%
+  mutate(ANRIND2 = paste(ANRIND, PARAMCD)) %>%
+  mutate(ANRIND3 = paste(ANRIND, PARAM))
+
+# Ensure character variables are converted to factors and empty strings and NAs are explicit missing levels.
+adsl_f <- df_explicit_na(adsl_f)
+adeg_f <- df_explicit_na(adeg_f)
+
+adeg_f_1 <- adeg_f %>%
+  filter(PARAMCD != "QT") %>%
+  mutate(CRITDIR = "Direction 1") %>%
+  mutate(ANRIND23 = ANRIND2)
+
+adeg_f_2 <- adeg_f %>%
+  mutate(CRITDIR = "Direction 2") %>%
+  mutate(ANRIND23 = ANRIND3)
+
+adeg_f <- rbind(adeg_f_1, adeg_f_2)
+
+
+
+### mapping dataframe
+uniq_adeg <- unique(adeg_f %>% select(PARCAT, PARAMCD, CRITDIR, ANRIND23))
+mapdf <- uniq_adeg %>%
+  arrange(PARCAT, PARAMCD, CRITDIR, ANRIND23) %>%
+  mutate(across(where(is.factor), as.character))
+
+### 3 versions of dataset
+### CRITDIR not factor, all visits have available data for all PARAMCD
+adeg_f1 <- adeg_f
+
+### CRITDIR not factor, not all visits available data for all PARAMCD
+### exclude QT from screening - this will generate the problem in 6.11
+adeg_f <- adeg_f %>%
+  filter(!(AVISIT == "SCREENING" & PARAMCD == "QT"))
+
+### CRITDIR factor, not all visits available data
+adeg_f2 <- adeg_f
+### also part of the issue
+adeg_f2$CRITDIR <- factor(adeg_f2$CRITDIR)
+
+
+
+
+
+
+################################################################################
+# Define layout and build table:
+################################################################################
+
+
+lyt <- basic_table(
+  show_colcounts = TRUE,
+  colcount_format = "N=xx"
+) %>%
+  split_cols_by("ACTARM") %>%
+  split_rows_by("AVISIT",
+    label_pos = "topleft",
+    section_div = " ",
+    split_fun = drop_split_levels,
+    split_label = "Time Point",
+    child_labels = "visible"
+  ) %>%
+  split_rows_by("PARAMCD",
+    split_label = "Test",
+    label_pos = "topleft",
+    indent_mod = 1L,
+    child_labels = "visible",
+    split_fun = trim_levels_to_map(mapdf)
+  ) %>%
+  split_rows_by("CRITDIR",
+    split_label = "Direction",
+    label_pos = "topleft",
+    indent_mod = 1L,
+    child_labels = "visible",
+    split_fun = trim_levels_to_map(mapdf)
+  ) %>%
+  analyze(c("ANRIND23"),
+    a_summary,
+    extra_args = list(.stats = "count_fraction"),
+    show_labels = "hidden",
+    indent_mod = 1L
+  )
+
+testthat::test_that("CRITDIR is factor, and one PARAMCD is not available on all AVISITS", {
+  res <- testthat::expect_silent(build_table(lyt, adeg_f2, alt_counts_df = adsl_f))
+  testthat::expect_snapshot(res)
+})
+
+
+testthat::test_that("CRITDIR is not factor, and all PARAMCD available on all AVISITS", {
+  res <- testthat::expect_silent(build_table(lyt, adeg_f1, alt_counts_df = adsl_f))
+  testthat::expect_snapshot(res)
+})
+
+testthat::test_that("CRITDIR is not factor, and one PARAMCD is not available on all AVISITS", {
+  res <- testthat::expect_silent(build_table(lyt, adeg_f, alt_counts_df = adsl_f))
+  testthat::expect_snapshot(res)
+})

tests/testthat/test-listing_adal02.R

@@ -20,7 +20,6 @@ testthat::test_that("ADAL02 listing is produced correctly", {
       values_from = AVAL
     ) %>%
     filter(if_any(matches("Treatment Emergent - Positive"), ~ .x == 1)) %>%
-    # filter(`Treatment Emergent - Positive` == 1) %>%
     mutate(
       VISN = factor(paste0(
         VISIT, "\n(Day ",

tests/testthat/test-table_aet04.R

@@ -25,11 +25,13 @@ adae <- level_reducer(adae, "AEDECOD", p_to_keep = 0.7)
 # Helper function to avoid filtering also the first part of the table, where general information is given.
 my_row_condition <- function(row_fnc_condition) {
   function(table_row) {
+    ret <- NULL
     if (indent_mod(table_row) == 0) {
-      return(TRUE)
+      ret <- TRUE
     } else {
-      row_fnc_condition(table_row)
+      ret <- row_fnc_condition(table_row)
     }
+    ret
   }
 }
 

tests/testthat/test-table_aovt01.R

@@ -51,3 +51,23 @@ testthat::test_that("AOVT01 variant with multiple endpoints is produced correctl
   res <- testthat::expect_silent(result)
   testthat::expect_snapshot(res)
 })
+
+testthat::test_that("AOVT01 variant with proportional weights is produced correctly", {
+  adqs_multi <- dplyr::filter(adqs, AVISIT == "WEEK 1 DAY 8")
+  n_per_arm <- table(adsl$ARM)
+
+  result <- basic_table() %>%
+    split_cols_by("ARMCD", ref_group = "ARM A", split_fun = ref_group_position("first")) %>%
+    add_colcounts() %>%
+    split_rows_by("PARAMCD") %>%
+    summarize_ancova(
+      vars = "CHG",
+      variables = list(arm = "ARMCD", covariates = c("BASE", "STRATA1")),
+      conf_level = 0.95, var_labels = "Adjusted mean",
+      weights_emmeans = "proportional"
+    ) %>%
+    build_table(adqs_multi, alt_counts_df = adsl)
+
+  res <- testthat::expect_silent(result)
+  testthat::expect_snapshot(res)
+})

tests/testthat/test-table_pkpt04.R

@@ -33,7 +33,6 @@ l <- basic_table() %>%
 testthat::test_that("PKPT04 is produced correctly for Drug X", {
   adpp0 <- adpp_urine %>%
     filter(PPCAT == "XANOMELINE") %>%
-    # h_pkparam_sort() %>%
     mutate(PKPARAM = factor(paste0(PPTEST, " (", AVALU, ")")))
 
   result <- build_table(l, df = adpp0)

tests/testthat/test-table_pkpt05.R

@@ -56,7 +56,6 @@ l <- basic_table() %>%
 testthat::test_that("PKPT05 Drug X is produced correctly", {
   adpp0 <- adpp_urine %>%
     filter(PPCAT == "XANOMELINE") %>%
-    # h_pkparam_sort() %>%
     dplyr::mutate(PKPARAM = factor(paste0(PPTEST, " (", AVALU, ")")))
   result <- build_table(l, df = adpp0)
   main_title(result) <- paste("Summary of", unique(adpp0$PPSPEC), "PK Parameter by Treatment Arm, PK Population")