get_dupes() grouped data error fix (#345)

* Add check for grouped data, if so remove groups, check for dupes, and reapply groups.

R/get_dupes.R

@@ -24,6 +24,18 @@ get_dupes <- function(dat, ...) {
 
   names(dat)[pos] <- names(pos) #allows for renaming within get_dupes() consistent with select()
 
+  #Check if dat is grouped and if so, save structure and ungroup temporarily
+  is_grouped <- dplyr::is_grouped_df(dat)
+
+  if(is_grouped) {
+    dat_groups <- dplyr::group_vars(dat)
+    dat <- dat %>% dplyr::ungroup()
+    if(getOption("get_dupes.grouped_warning",TRUE) & interactive()) {
+      message(paste0("Data is grouped by [", paste(dat_groups, collapse = "|"), "]. Note that get_dupes() is not group aware and does not limit duplicate detection to within-groups, but rather checks over the entire data frame. However grouping structure is preserved.\nThis message is shown once per session and may be disabled by setting options(\"get_dupes.grouped_warning\" = FALSE).")) #nocov
+      options("get_dupes.grouped_warning" = FALSE) #nocov
+    }
+  }
+
   if (rlang::dots_n(...) == 0) { # if no tidyselect variables are specified, check the whole data.frame
     var_names <- names(dat)
     nms <- rlang::syms(var_names)
@@ -54,7 +66,11 @@ get_dupes <- function(dat, ...) {
   if (nrow(dupes) == 0) {
     message(paste0("No duplicate combinations found of: ", paste(var_names, collapse = ", ")))
   }
-  dupes
+
+  #Reapply groups if dat was grouped
+  if(is_grouped) dupes <- dupes %>% dplyr::group_by(!!!rlang::syms(dat_groups))
+
+  return(dupes)
 }
 
 

tests/testthat/test-get-dupes.R

@@ -43,3 +43,10 @@ test_that("tidyselect specification matches exact specification", {
   expect_equal(mtcars %>% get_dupes(mpg), mtcars %>% get_dupes(-c(cyl, disp, hp, drat, wt, qsec, vs, am ,gear, carb)))
   expect_equal(suppressMessages(mtcars %>% select(cyl, wt) %>% get_dupes()), mtcars %>% select(cyl, wt) %>% get_dupes(everything()))
 })
+
+test_that("grouped and ungrouped data is handled correctly", {
+  expect_equal(suppressMessages(mtcars %>% group_by(carb, cyl) %>% get_dupes(mpg, carb)) %>% group_vars(), 
+               mtcars %>% group_by(carb, cyl) %>% group_vars())
+  expect_equal(suppressMessages(mtcars %>% group_by(carb, cyl) %>% get_dupes(mpg, carb) %>% ungroup()),
+               mtcars %>% get_dupes(mpg, carb))
+})