test_that("select preserves grouping", { gf <- duckplyr_group_by(tibble(g = 1:3, x = 3:1), g) i <- count_regroups(out <- duckplyr_select(gf, h = g)) expect_equal(i, 0) expect_equal(duckplyr_group_vars(out), "h") }) test_that("grouping variables preserved with a message, unless already selected (#1511, #5841)", { df <- tibble(g = 1:3, x = 3:1) %>% duckplyr_group_by(g) expect_snapshot({ res <- duckplyr_select(df, x) }) expect_named(res, c("g", "x")) df <- tibble(a = 1, b = 2, c = 3) %>% duckplyr_group_by(a) expect_equal(df %>% duckplyr_select(a = b), tibble(a = 2)) df <- tibble(a = 1, b = 2, c = 3) %>% duckplyr_group_by(a, b) expect_snapshot({ expect_equal(df %>% duckplyr_select(a = c), tibble(b = 2, a = 3) %>% duckplyr_group_by(b)) expect_equal(df %>% duckplyr_select(b = c), tibble(a = 1, b = 3) %>% duckplyr_group_by(a)) }) }) test_that("non-syntactic grouping variable is preserved (#1138)", { expect_snapshot( df <- tibble(`a b` = 1L) %>% duckplyr_group_by(`a b`) %>% duckplyr_select() ) expect_named(df, "a b") }) test_that("select doesn't fail if some names missing", { df1 <- data.frame(x = 1:10, y = 1:10, z = 1:10) df2 <- setNames(df1, c("x", "y", "")) # df3 <- setNames(df1, c("x", "", "")) expect_equal(duckplyr_select(df1, x), data.frame(x = 1:10)) expect_equal(duckplyr_select(df2, x), data.frame(x = 1:10)) # expect_equal(duckplyr_select(df3, x), data.frame(x = 1:10)) }) # Special cases ------------------------------------------------- test_that("select with no args returns nothing", { empty <- duckplyr_select(mtcars) expect_equal(ncol(empty), 0) expect_equal(nrow(empty), 32) empty <- duckplyr_select(mtcars, !!!list()) expect_equal(ncol(empty), 0) expect_equal(nrow(empty), 32) }) test_that("select excluding all vars returns nothing", { expect_equal(dim(duckplyr_select(mtcars, -(mpg:carb))), c(32, 0)) expect_equal(dim(duckplyr_select(mtcars, starts_with("x"))), c(32, 0)) expect_equal(dim(duckplyr_select(mtcars, -matches("."))), c(32, 0)) }) test_that("negating empty match returns everything", { df <- data.frame(x = 1:3, y = 3:1) expect_equal(duckplyr_select(df, -starts_with("xyz")), df) }) test_that("can select with duplicate columns", { skip_if(Sys.getenv("DUCKPLYR_FORCE") == "TRUE") df <- tibble(x = 1, x = 2, y = 1, .name_repair = "minimal") # can extract duplicate cols by position expect_named(df %>% duckplyr_select(1, 3), c("x", "y")) # can select out non-duplicated columns expect_named(df %>% duckplyr_select(y), "y") }) # Select variables ----------------------------------------------- test_that("select can be before group_by (#309)", { df <- data.frame( id = c(1, 1, 2, 2, 2, 3, 3, 4, 4, 5), year = c(2013, 2013, 2012, 2013, 2013, 2013, 2012, 2012, 2013, 2013), var1 = rnorm(10) ) dfagg <- df %>% duckplyr_group_by(id, year) %>% duckplyr_select(id, year, var1) %>% duckplyr_summarise(var1 = mean(var1)) expect_equal(names(dfagg), c("id", "year", "var1")) }) test_that("select succeeds in presence of raw columns (#1803)", { skip_if(Sys.getenv("DUCKPLYR_FORCE") == "TRUE") df <- tibble(a = 1:3, b = as.raw(1:3)) expect_identical(duckplyr_select(df, a), df["a"]) expect_identical(duckplyr_select(df, b), df["b"]) expect_identical(duckplyr_select(df, -b), df["a"]) }) test_that("arguments to duckplyr_select() don't match vars_select() arguments", { df <- tibble(a = 1) expect_identical(duckplyr_select(df, var = a), tibble(var = 1)) expect_identical(duckplyr_select(duckplyr_group_by(df, a), var = a), duckplyr_group_by(tibble(var = 1), var)) expect_identical(duckplyr_select(df, exclude = a), tibble(exclude = 1)) expect_identical(duckplyr_select(df, include = a), tibble(include = 1)) expect_identical(duckplyr_select(duckplyr_group_by(df, a), exclude = a), duckplyr_group_by(tibble(exclude = 1), exclude)) expect_identical(duckplyr_select(duckplyr_group_by(df, a), include = a), duckplyr_group_by(tibble(include = 1), include)) }) test_that("can duckplyr_select() with deprecated `.data` pronoun (#2715)", { withr::local_options(lifecycle_verbosity = "quiet") expect_identical(duckplyr_select(mtcars, .data$cyl), duckplyr_select(mtcars, cyl)) }) test_that("can duckplyr_select() with character vectors", { expect_identical(duckplyr_select(mtcars, "cyl", !!"disp", c("cyl", "am", "drat")), mtcars[c("cyl", "disp", "am", "drat")]) }) test_that("duckplyr_select() treats NULL inputs as empty", { expect_identical(duckplyr_select(mtcars, cyl), duckplyr_select(mtcars, NULL, cyl, NULL)) }) test_that("can duckplyr_select() with strings and character vectors", { vars <- c(foo = "cyl", bar = "am") expect_identical(duckplyr_select(mtcars, !!!vars), duckplyr_select(mtcars, foo = cyl, bar = am)) expect_identical(duckplyr_select(mtcars, !!vars), duckplyr_select(mtcars, foo = cyl, bar = am)) }) test_that("select works on empty names (#3601)", { df <- data.frame(x=1, y=2, z=3) colnames(df) <- c("x","y","") expect_identical(duckplyr_select(df, x)$x, 1) colnames(df) <- c("","y","z") expect_identical(duckplyr_select(df, y)$y, 2) }) test_that("select works on NA names (#3601)", { df <- data.frame(x=1, y=2, z=3) colnames(df) <- c("x","y",NA) expect_identical(duckplyr_select(df, x)$x, 1) colnames(df) <- c(NA,"y","z") expect_identical(duckplyr_select(df, y)$y, 2) }) test_that("duckplyr_select() keeps attributes of raw data frames (#5831)", { df <- data.frame(x = 1) attr(df, "a") <- "b" expect_equal(attr(duckplyr_select(df, x), "a"), "b") }) test_that("duckplyr_select() provides informative errors", { expect_snapshot({ (expect_error(duckplyr_select(mtcars, 1 + ""))) }) }) # dplyr_col_select() ------------------------------------------------------ test_that("dplyr_col_select() aborts when `[` implementation is broken", { local_methods( "[.dplyr_test_broken_operator" = function(x, ...) { unclass(x) }, "[.dplyr_test_operator_wrong_size" = function(x, ...) { data.frame() } ) df1 <- new_tibble(list(x = 1), nrow = 1L, class = "dplyr_test_broken_operator") expect_snapshot({ (expect_error( duckplyr_select(df1, 1:2) )) (expect_error( duckplyr_select(df1, 0) )) }) df2 <- new_tibble(list(x = 1), nrow = 1L, class = "dplyr_test_operator_wrong_size") expect_error(duckplyr_select(df2, 1:2)) expect_snapshot({ # from vctrs (expect_error( duckplyr_select(df1, 2) )) # not returning a data frame (expect_error( duckplyr_select(df1, 1) )) # unexpected number of columns (expect_error( duckplyr_select(df2, 1) )) }) })