Update in tbl_ard_summary() to assign default summary types for non-standard ARDs

DESCRIPTION

@@ -80,6 +80,7 @@ Suggests:
     parsnip (>= 0.1.7),
     rmarkdown,
     smd (>= 0.6.6),
+    spelling,
     survey (>= 4.2),
     survival (>= 3.6-4),
     testthat (>= 3.2.0),

NEWS.md

@@ -1,5 +1,7 @@
 # gtsummary (development version)
 
+* Update in `tbl_ard_summary()` to better handle non-standard ARDs (i.e. not our typical continuous or categorical summaries) by assigning them a default summary type. (#1991)
+
 * Made the `oneway.test()` available in `add_p.tbl_continuous()`. (#1970)
 
 * Added argument `tbl_ard_summary(overall)`. When `TRUE`, the ARD is parsed into primary ARD and the Overall ARD and we run `tbl_ard_summary() |> add_overall()`. (#1940)
@@ -138,7 +140,7 @@ Updates to address regressions in the v2.0.0 release:
 
 * Added `pkgdown_print.gtsummary()` method that is only registered when the pkgdown package is loaded. This enables printing of gtsummary tables on the pkgdown site in the Examples section. (#1771)
 
-* The package now uses updated `survey::svyquantile()` function to calculate quatiles, which was introduced in survey v4.1
+* The package now uses updated `survey::svyquantile()` function to calculate quantiles, which was introduced in survey v4.1
 
 ### Bug fixes
 
@@ -170,7 +172,7 @@ Updates to address regressions in the v2.0.0 release:
 
 * Arguments `modify_header(update)`, `modify_footnote(update)`, `modify_spanning_header(update)`, and `modify_fmt_fun(update)` have been deprecated. Use dynamic dots instead, e.g. `modify_header(...)`, which has been the preferred method for passing updates for a few years.
 
-* Function `continuous_summary()` has been deprecated immediately. Apologies for the inconvenience of the immeidate deprecation. The way the function originally worked is not compatible with the updated internal structures. In most cases, users can use the `tbl_continuous()` function instead.
+* Function `continuous_summary()` has been deprecated immediately. Apologies for the inconvenience of the immediate deprecation. The way the function originally worked is not compatible with the updated internal structures. In most cases, users can use the `tbl_continuous()` function instead.
 
 * Arguments `add_stat(fmt_fun, header, footnote, new_col_name)` have been deprecated since v1.4.0 (2021-04-13). They have now been fully removed from the package.
 

R/add_p.R

@@ -56,7 +56,7 @@ add_p <- function(x, ...) {
 #' @section test argument:
 #'
 #' See the [?tests][tests] help file for details on available tests and creating custom tests.
-#' The [?tests][tests] help file also includes psuedo-code for each test to be clear
+#' The [?tests][tests] help file also includes pseudo-code for each test to be clear
 #' precisely how the calculation is performed.
 #'
 #' The default test used in `add_p()` primarily depends on these factors:

R/assign_tests.R

@@ -22,7 +22,7 @@
 #' @param cont_variable (`string`)\cr
 #'   a column name of the continuous summary variable in `tbl_continuous()`
 #' @param summary_type (named `list`)\cr
-#'   naemd list of summary types
+#'   named list of summary types
 #' @inheritParams cli::cli_abort
 #'
 #' @return A table of class `'gtsummary'`

R/tbl_ard_summary.R

@@ -165,6 +165,15 @@ tbl_ard_summary <- function(cards,
     ) |>
     deframe() |>
     as.list()
+  # for non-standard ARDs, fill in the missing default types
+  for (v in setdiff(include, names(default_types))) {
+    if (!"variable_level" %in% names(cards) ||
+        is_empty(compact(dplyr::filter(cards, .data$variable %in%.env$v)$variable_level))) {
+      default_types[[v]] <- "continuous"
+    }
+    else default_types[[v]] <- "categorical" # styler: off
+  }
+  if (exists("v")) remove("v") # styler: off
 
   # process arguments ----------------------------------------------------------
   cards::process_formula_selectors(
@@ -187,14 +196,14 @@ tbl_ard_summary <- function(cards,
     walk(
       include,
       function(variable) {
-        if (default_types[[variable]] %in% "continuous" &&
+        if (isTRUE(default_types[[variable]] %in% "continuous") &&
             !type[[variable]] %in% c("continuous", "continuous2")) {
           cli::cli_abort(
             "Summary type for variable {.val {variable}} must be one of
              {.val {c('continuous', 'continuous2')}}, not {.val {type[[variable]]}}.",
             call = get_cli_abort_call()
           )
-        } else if (default_types[[variable]] %in% c("categorical", "dichotomous") &&
+        } else if (isTRUE(default_types[[variable]] %in% c("categorical", "dichotomous")) &&
                    !identical(type[[variable]], default_types[[variable]])) {
           cli::cli_abort(
             "Summary type for variable {.val {variable}} must be

R/tbl_uvregression.R

@@ -40,7 +40,7 @@
 #'   is the dependent variable, and `{x}` represents a single covariate. For a
 #'   random intercept model, the formula may be `formula = "{y} ~ {x} + (1 | gear)"`.
 #' @param method.args (named `list`)\cr
-#'   Named list of arguments assed to `method`.
+#'   Named list of arguments passed to `method`.
 #' @param hide_n (scalar `logical`)\cr
 #'   Hide N column. Default is `FALSE`
 #' @inheritParams tbl_regression

R/theme_gtsummary.R

@@ -8,7 +8,7 @@
 #' @param set_theme (scalar `logical`)\cr
 #'   Logical indicating whether to set the theme. Default is `TRUE`.
 #'   When `FALSE` the named list of theme elements is returned invisibly
-#' @param font_size (scaler `numeric`)\cr
+#' @param font_size (scalar `numeric`)\cr
 #'   Numeric font size for compact theme.
 #'   Default is 13 for gt tables, and 8 for all other output types
 #'
@@ -29,7 +29,7 @@
 #'       - `tbl_summary()` all percentages rounded to one decimal place
 #'       - `tbl_regression()`,`tbl_uvregression()` add significance stars with `add_significance_stars()`;
 #'          hides CI and p-value from output
-#'           - For flaxtable and huxtable output, the coeficient's standard error is placed below. For gt, it is placed to the right.
+#'           - For flextable and huxtable output, the coefficients' standard error is placed below. For gt, it is placed to the right.
 #' - `theme_gtsummary_compact()`
 #'   - tables printed with gt, flextable, kableExtra, or huxtable will be compact with smaller font size and reduced cell padding
 #' - `theme_gtsummary_printer(print_engine)`

inst/WORDLIST

@@ -1,14 +1,23 @@
 ANCOVA
+ARD
+ARDs
+Agresti
+BH
 Biostatistics
+CDISC
 CMD
 Codecov
+Coull
 DOI
 GGally
 GVIF
 Gehan
 Hmisc
 JAMA
+Jeffreys
 Kaplan
+Lifecycle
+Likert
 MacOS
 McNemar
 McNemar's
@@ -21,20 +30,22 @@ RJ
 RStudio
 RTF
 Rao
+Rmarkdown
 SHA
 Tarone
 Tidiers
+Univariable
 VIF
 Waerden's
 Wainberg
 YAML
+bonferroni
 bstfun
+cardx
 cli
 codebase
 coef
 conf
-coxph
-crosstab
 customizability
 customizable
 der
@@ -44,27 +55,30 @@ dplyr
 effectsize
 emmeans
 exponentiate
+fdr
 flextable
 forcats
 ftExtra
 ggplot
 ggstats
-glm
+hochberg
+holm
+hommel
 huxtable
 kable
 kableExtra
 knitr
 labelled
 lifecycle
-lm
+likert
+linebreaks
 lme
-logLik
 mL
-mira
+mis
 nevent
 ng
 nnet
-obejcts
+pkgdown
 pre
 pvalue
 quosure
@@ -80,11 +94,14 @@ th
 tibble
 tibbles
 tidiers
+tidycmprsk
 tidyr
 tidyselect
 tidyselect's
 tidyverse
+tilda
 un
+unhidden
 unhide
 unicode
 univariable

tests/spelling.R

@@ -0,0 +1,3 @@
+if(requireNamespace('spelling', quietly = TRUE))
+  spelling::spell_check_test(vignettes = TRUE, error = FALSE,
+                             skip_on_cran = TRUE)

tests/testthat/_snaps/tbl_ard_summary.md

@@ -163,3 +163,14 @@
       Error in `tbl_ard_summary()`:
       ! Variable "AGE" is type `continuous` and `statistic` argument value must be a string of length one.
 
+# tbl_ard_summary() non-standard ARDs (ie not 'continuous', 'categorical', etc)
+
+    Code
+      as.data.frame(tbl_ard_summary(ard, by = trt, statistic = ~"{estimate}"))
+    Output
+        **Characteristic** **Drug A** **Drug B**
+      1               time       <NA>       <NA>
+      2                 12       90.8       86.3
+      3                 24       46.9       41.2
+      4                age       47.0       47.4
+

tests/testthat/test-tbl_ard_summary.R

@@ -297,3 +297,40 @@ test_that("tbl_ard_summary() existing 'gts_column'", {
   )
 })
 
+test_that("tbl_ard_summary() non-standard ARDs (ie not 'continuous', 'categorical', etc)", {
+  # This ARD was created with this code:
+  # cards::bind_ard(
+  #   survival::survfit(survival::Surv(ttdeath, death) ~ trt, trial) |>
+  #     cardx::ard_survival_survfit(times = c(12, 24)) |>
+  #     dplyr::filter(stat_name %in% c("estimate")) |>
+  #     dplyr::mutate(
+  #       fmt_fn = list("xx.x%"),
+  #       group1_level = unlist(group1_level) |> as.character() |> as.list()
+  #     ),
+  #   cardx::ard_stats_t_test_onesample(trial, variables = age, by = trt) |>
+  #     dplyr::filter(stat_name %in% c("estimate"))
+  # ) |>
+  #   dplyr::select(-cards::all_missing_columns()) |>
+  #   dput()
+
+  ard <-
+    structure(list(group1 = c("trt", "trt", "trt", "trt", "trt",
+                              "trt"), group1_level = list("Drug A", "Drug A", "Drug B", "Drug B",
+                                                          "Drug A", "Drug B"), variable = c("time", "time", "time",
+                                                                                            "time", "age", "age"), variable_level = list(12, 24, 12, 24,
+                                                                                                                                         NULL, NULL), context = c("survival", "survival", "survival",
+                                                                                                                                                                  "survival", "stats_t_test_onesample", "stats_t_test_onesample"
+                                                                                                                                         ), stat_name = c("estimate", "estimate", "estimate", "estimate",
+                                                                                                                                                          "estimate", "estimate"), stat_label = c("Survival Probability",
+                                                                                                                                                                                                  "Survival Probability", "Survival Probability", "Survival Probability",
+                                                                                                                                                                                                  "Mean", "Mean"), stat = list(0.908163265306122, 0.46938775510204,
+                                                                                                                                                                                                                               0.862745098039216, 0.411764705882353, c(`mean of x` = 47.010989010989),
+                                                                                                                                                                                                                               c(`mean of x` = 47.4489795918367)), fmt_fn = list("xx.x%",
+                                                                                                                                                                                                                                                                                 "xx.x%", "xx.x%", "xx.x%", 1L, 1L)), row.names = c(NA, -6L
+                                                                                                                                                                                                                                                                                 ), class = c("card", "tbl_df", "tbl", "data.frame"))
+
+  expect_snapshot(
+    tbl_ard_summary(ard, by = trt, statistic = ~ "{estimate}") |>
+      as.data.frame()
+  )
+})