require optional arguments to be named

DESCRIPTION

@@ -19,7 +19,7 @@ URL: https://github.com/tidymodels/finetune,
 BugReports: https://github.com/tidymodels/finetune/issues
 Depends: 
     R (>= 3.5),
-    tune (>= 1.1.2.9011)
+    tune (>= 1.1.2.9020)
 Imports: 
     cli,
     dials (>= 0.1.0),

NEWS.md

@@ -10,12 +10,11 @@
 
 * Enabling the `verbose_elim` control option for `tune_race_anova()` will now additionally introduce a message confirming that the function is evaluating against the burn-in resamples.
 
-* Updates based on the new version of tune, primarily for survival analysis models. 
+* Updates based on the new version of tune, primarily for survival analysis models (#104). 
 
 ## Breaking Change
 
-* `show_best.tune_race()` gains an `eval_time` argument for censored regression models. This breaks passing `n` by position (#104).
-
+* Ellipses (...) are now used consistently in the package to require optional arguments to be named. `collect_predictions()`, `collect_metrics()` and `show_best()` methods previously had ellipses at the end of the function signature  that have been moved to follow the last argument without a default value. Optional arguments previously passed by position will now error informatively prompting them to be named (#105).
 
 # finetune 1.1.0
 

R/racing_helpers.R

@@ -665,12 +665,13 @@ randomize_resamples <- function(x) {
 #' @name collect_predictions
 collect_predictions.tune_race <-
   function(x,
+           ...,
            summarize = FALSE,
            parameters = NULL,
-           all_configs = FALSE,
-           ...) {
+           all_configs = FALSE) {
+    rlang::check_dots_empty()
     x <- dplyr::select(x, -.order)
-    res <- NextMethod(summarize = summarize, parameters = parameters)
+    res <- collect_predictions(x, summarize = summarize, parameters = parameters)
     if (!all_configs) {
       final_configs <- subset_finished_race(x)
       res <- dplyr::inner_join(res, final_configs, by = ".config")
@@ -681,10 +682,11 @@ collect_predictions.tune_race <-
 #' @inheritParams tune::collect_metrics
 #' @export
 #' @rdname collect_predictions
-collect_metrics.tune_race <- function(x, summarize = TRUE, all_configs = FALSE, ...) {
+collect_metrics.tune_race <- function(x, ..., summarize = TRUE, type = c("long", "wide"), all_configs = FALSE) {
+  rlang::check_dots_empty()
   x <- dplyr::select(x, -.order)
   final_configs <- subset_finished_race(x)
-  res <- NextMethod(summarize = summarize, ...)
+  res <- collect_metrics(x, summarize = summarize, type = type)
   if (!all_configs) {
     final_configs <- subset_finished_race(x)
     res <- dplyr::inner_join(res, final_configs, by = ".config")
@@ -704,7 +706,13 @@ collect_metrics.tune_race <- function(x, summarize = TRUE, all_configs = FALSE,
 #' resampled). Comparing performance metrics for configurations averaged with
 #' different resamples is likely to lead to inappropriate results.
 #' @export
-show_best.tune_race <- function(x, metric = NULL, eval_time = NULL, n = 5, ...) {
+show_best.tune_race <- function(x,
+                                ...,
+                                metric = NULL,
+                                eval_time = NULL,
+                                n = 5,
+                                call = rlang::current_env()) {
+  rlang::check_dots_empty()
   if (!is.null(metric)) {
     # What was used to judge the race and how are they being sorted now?
     metrics <- tune::.get_tune_metrics(x)
@@ -722,7 +730,7 @@ show_best.tune_race <- function(x, metric = NULL, eval_time = NULL, n = 5, ...)
   x <- dplyr::select(x, -.order)
   final_configs <- subset_finished_race(x)
 
-  res <- NextMethod(metric = metric, eval_time = eval_time, n = Inf, ...)
+  res <- NextMethod(metric = metric, eval_time = eval_time, n = Inf, call = call)
   res$.ranked <- 1:nrow(res)
   res <- dplyr::inner_join(res, final_configs, by = ".config")
   res$.ranked <- NULL

R/tune_race_anova.R

@@ -303,8 +303,8 @@ tune_race_anova_workflow <-
 
     opt_metric_time <- tune::first_eval_time(
       metrics,
-      opt_metric_name,
-      eval_time,
+      metric = opt_metric_name,
+      eval_time = eval_time,
       call = call
     )
 

R/tune_race_win_loss.R

@@ -283,8 +283,8 @@ tune_race_win_loss_workflow <-
 
     opt_metric_time <- tune::first_eval_time(
       metrics,
-      opt_metric_name,
-      eval_time,
+      metric = opt_metric_name,
+      eval_time = eval_time,
       call = call
     )
 

R/tune_sim_anneal.R

@@ -344,8 +344,8 @@ tune_sim_anneal_workflow <-
     eval_time <- tune::check_eval_time_arg(eval_time, metrics, call = call)
     opt_metric_time <- tune::first_eval_time(
         metrics,
-        opt_metric_name,
-        eval_time,
+        metric = opt_metric_name,
+        eval_time = eval_time,
         call = call
       )