Merge pull request #19 from poissonconsulting/duplicate2

- Added `.keep_all` argument to `duplicates()`

R/duplicates.R

@@ -1,16 +1,24 @@
-#' Find duplicate rows in a data frame
+#' Keep non-unique rows in a data frame
 #'
-#' Finds duplicate rows within a data frame. Column names can optionally be
-#' provided. If column names aren't provided, the entire data frame (all
-#' columns) will be searched for duplicate rows. If column names are provided,
-#' only the specified columns will be searched for duplicate rows.
+#' Keeps only non-unique rows within a data frame.
 #' 
 #' @param .data A data.frame.
-#' @param ... Optional column names to use when searching for duplicate rows in specific columns.
-#' @return A tibble containing only duplicate rows.
+#' @param ... Optional variables to use when determining non-uniqueness.
+#'  If omitted, will use all variables in the data frame.
+#' @param .keep_all A flag specifying whether to keep all variables in .data.
+#' @return The original data frame with only non-unique rows.
 #' @export
-duplicates <- function(.data, ...) {
+#' @examples 
+#' data <- tibble::tibble(x = c(1,2,1,1), y = c(1,1,1,5))
+#' 
+#' duplicates(data)
+#' duplicates(data, x)
+#' duplicates(data, y)
+#' duplicates(data, x, y)
+#' duplicates(data, y, .keep_all = FALSE)
+duplicates <- function(.data, ..., .keep_all = TRUE) {
   check_data(.data)
+  chk_flag(.keep_all)
 
   col <- rlang::ensyms(...)
   if (length(col) == 0) {
@@ -27,12 +35,12 @@ duplicates <- function(.data, ...) {
     return(.data)
   }
   .data_dup <- dplyr::select(.data, dplyr::all_of(col_names))
-  print(.data_dup)
   .data_dup <- .data_dup[duplicated(.data_dup), , drop = FALSE]
-  print(.data_dup)
   .data_dup <- unique(.data_dup)
-  print(.data_dup)
-  .data <- merge(.data, .data_dup, by = col_names)
+  .data <- dplyr::inner_join(.data, .data_dup, by = col_names)
+  if(!(.keep_all)) {
+    .data <- dplyr::select(.data, dplyr::all_of(col_names))
+  }
   .data <- dplyr::as_tibble(.data)
   .data
 }

tests/testthat/test-duplicates.R

@@ -3,23 +3,89 @@ test_that("returns only duplicated rows of selected columns", {
   expect_identical(duplicates(tib), tib[c(1, 3), ])
   expect_identical(duplicates(as.data.frame(tib)), tib[c(1, 3), ])
 
-  expect_equal(
+  expect_identical(
     duplicates(data.frame(x = c(1, 2, 1), y = 1:3), x),
-    dplyr::tibble(x = c(1, 1), y = c(1, 3))
+    dplyr::tibble(x = c(1, 1), y = c(1L, 3L))
   )
 
-  expect_equal(
+  expect_identical(
     duplicates(data.frame(x = c(1, 2, 1), y = 1:3), x, y),
-    dplyr::tibble(x = double(0), y = double(0))
+    dplyr::tibble(x = double(0), y = integer(0))
   )
 })
 
-test_that("errors when no input argument is supplied", {
-  expect_error(
-    duplicates(),
-    'argument ".data" is missing, with no default',
-    fixed = TRUE
+test_that("keep_all working", {
+  data <- tibble::tibble(x = c(1,2,1,1), y = c(1,1,1,5))
+  expect_identical(duplicates(data, y),
+                   tibble::tibble(x = c(1,2,1), y = c(1,1,1)))
+  expect_identical(duplicates(data, x, y),
+                   tibble::tibble(x = c(1,1), y = c(1,1)))
+  expect_identical(duplicates(data, y, x),
+                   tibble::tibble(x = c(1,1), y = c(1,1)))
+  expect_identical(duplicates(data),
+                   tibble::tibble(x = c(1,1), y = c(1,1)))
+  expect_identical(duplicates(data, y, .keep_all = FALSE),
+                   tibble::tibble(y = c(1,1,1)))
+})
+
+
+test_that("handles data frame with no rows", {
+  data <- dplyr::tibble(x = integer(), y = integer())
+
+  expect_equal(
+    duplicates(data),
+    data
   )
+
+  expect_equal(
+    duplicates(data, x),
+    data
+  )
+
+  expect_equal(
+    duplicates(data, x, .keep_all = FALSE),
+    dplyr::tibble(x = integer())
+  )
+})
+
+test_that("handles data frame with no columns", {
+  data <- dplyr::tibble()
+
+  expect_identical(
+    duplicates(data),
+    data
+  )
+
+  expect_identical(
+    duplicates(data, .keep_all = FALSE),
+    data
+  )
+})
+
+test_that("handles columns with missing values", {
+  data <- tibble::tibble(x = c(1,2,NA,1,1), y = c(1,1,NA,NA,NA))
+
+  expect_identical(
+    duplicates(data),
+    tibble::tibble(x = c(1, 1), y = as.double(c(NA, NA)))
+  )
+
+  expect_identical(
+    duplicates(data, y),
+    data
+  )
+
+  expect_identical(
+    duplicates(data, y, .keep_all = FALSE),
+    tibble::tibble(y = c(1, 1, NA, NA, NA))
+  )
+})
+
+test_that("handles data set with no duplicates", {
+  data <- tibble::tibble(x = c(1,2,NA), z = 1:3)
+  expect_identical(duplicates(data), tibble::tibble(x = double(), z = integer()))
+  expect_identical(duplicates(data, x), tibble::tibble(x = double(), z = integer()))
+  expect_identical(duplicates(data, x, .keep_all = FALSE), tibble::tibble(x = double()))
 })
 
 test_that("errors when input argument is not a data.frame", {