# Set number of data.table threads to 2
data.table::setDTthreads(threads = 2L)
# Set number of collapse threads to 1
collapse::set_collapse(nthreads = 1L)

testthat::test_that("stat_summarise", {
  iris2 <- dplyr::as_tibble(fslice_sample(iris, n = 10^3, replace = TRUE, seed = 918291))

  stat_summarise(fslice(iris2, 0))
  stat_summarise(fslice(iris2, 0), stat = "mean")
  stat_summarise(fslice(data.table::as.data.table(iris2), 0), stat = "mean")
  stat_summarise(fslice(data.table::as.data.table(iris2), 0), stat = "n")
  stat_summarise(fslice(data.table::as.data.table(iris2), 0), stat = c("n", "min"))

  stat_summarise(fslice(iris2, 0), stat = .stat_fns)

  stat_summarise(iris2)
  stat_summarise(iris2, stat = "mean",
                 .cols = "Sepal.Width")
  stat_summarise(iris2, stat = "mean",
                 .cols = "Sepal.Width",
                 .by = dplyr::all_of("Species"), sort = TRUE)
  stat_summarise(iris2, stat = "mean",
                 .cols = "Sepal.Width",
                 .by = dplyr::all_of("Species"), sort = FALSE)
  stat_summarise(iris2, stat = c("min", "max"),
                 .cols = 1)
  stat_summarise(iris2, stat = c("min", "max"),
                 .cols = 1, .names = "{.fn}")
  stat_summarise(iris2, stat = c("min", "mean", "max"),
                 .cols = 1:2)
  testthat::expect_error(stat_summarise(iris2, stat = c("min", "mean", "max"),
                 .cols = 1:2, .names = "{.fn}"))
  testthat::expect_error(stat_summarise(iris2, stat = c("min", "mean", "max"),
                 .cols = 1:2, .names = "{.col}"))
  testthat::expect_error(stat_summarise(iris2, stat = c("min", "max"),
                 .cols = 1, .names = "{.col}"))
  testthat::expect_equal(
    iris2 %>%
      stat_summarise(Species, stat = "first", .by = Species) %>%
      names(),
    c("Species", "first")
    )

  testthat::expect_equal(
    iris2 %>%
      stat_summarise(Species, stat = "first", .by = Species,
                     q_probs = c(0, 1)) %>%
      names(),
    c("Species", "first")
    )
  testthat::expect_equal(
    iris2 %>%
      stat_summarise(Species, Sepal.Length, stat = "first", .by = Species,
                     q_probs = c(0, 0.0005, 1)) %>%
      names(),
    c("Species", "Sepal.Length", "p0", "p0.05", "p100")
  )
  testthat::expect_equal(
    iris2 %>%
      stat_summarise(Species, Sepal.Length, stat = "first", .by = Species,
                     q_probs = c(0, 0.0005, 1), .names = "{.col}_first") %>%
      names(),
    c("Species", "Sepal.Length_first", "p0", "p0.05", "p100")
    # c("Species", "Species_first", "Sepal.Length_first", "p0", "p0.05", "p100")
  )
  mean_sl <- mean(iris2$Sepal.Length)
  testthat::expect_equal(iris2 %>%
                           transmute2(mean = mean(Sepal.Length)) %>%
                           stat_summarise(mean, stat = "mean"),
                         data.table::data.table(mean = mean_sl))
  testthat::expect_equal(iris2 %>%
                           transmute2(mean = mean(Sepal.Length)) %>%
                           stat_summarise(mean, stat = "mean", .names = "ok"),
                         data.table::data.table(ok = mean_sl))
  testthat::expect_equal(iris2 %>%
                           transmute2(mean = mean(Sepal.Length)) %>%
                           stat_summarise(.cols = c("ok" = "mean"), stat = "mean"),
                         data.table::data.table(mean = mean_sl))
  testthat::expect_equal(iris2 %>%
                           transmute2(mean = mean(Sepal.Length)) %>%
                           stat_summarise(.cols = c("ok" = "mean"), stat = "mean", .names = "{.col}"),
                         data.table::data.table(ok = mean_sl))
  testthat::expect_equal(iris2 %>%
                           transmute2(mean = mean(Sepal.Length)) %>%
                           stat_summarise(mean, stat = "max"),
                         data.table::data.table(max = mean_sl))
  testthat::expect_equal(iris2 %>%
                           transmute2(mean = mean(Sepal.Length)) %>%
                           stat_summarise(mean, stat = "mean", .by = mean),
                         data.table::data.table(mean = mean_sl))
  testthat::expect_equal(iris2 %>%
                           transmute2(mean2 = mean(Sepal.Length)) %>%
                           stat_summarise(mean2, stat = "mean", .by = mean2),
                         data.table::data.table(mean2 = mean_sl,
                                                mean = mean_sl))
  # iris2 %>%
  #   transmute2(mean = mean(Sepal.Length)) %>%
  #   stat_summarise(mean, stat = "max", .by = mean)
  # iris2 %>%
  #   transmute2(mean = mean(Sepal.Length)) %>%
  #   stat_summarise(mean, stat = "max", .by = mean, .names = "{.fn}")

  # Can't rename groups through tidy-select in this way
  testthat::expect_equal(
    iris2 %>%
    transmute2(mean = mean(Sepal.Length)) %>%
    stat_summarise(.cols = c("ok" = "mean"), stat = "max",
                   .by = mean, .names = "ok"),
    data.table::data.table(ok = mean_sl,
                           ok = mean_sl)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length)) %>%
      stat_summarise(ok = mean, stat = "max",
                     .by = mean, .names = "ok"),
    data.table::data.table(mean = mean_sl,
                           ok = mean_sl)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length)) %>%
      stat_summarise(ok = mean, stat = "max", .by = mean),
    data.table::data.table(mean = mean_sl,
                           max = mean_sl)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length)) %>%
      stat_summarise(ok = mean, stat = "max", .by = mean, .names = "{.col}_max"),
    data.table::data.table(mean = mean_sl,
                           ok_max = mean_sl)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length)) %>%
      stat_summarise(yeah = mean, stat = "max", .by = mean, .names = "ok"),
    data.table::data.table(mean = mean_sl,
                           ok = mean_sl)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length)) %>%
      stat_summarise(.cols = c("ok" = "mean"), stat = "max", .names = "ok"),
    data.table::data.table(ok = mean_sl)
  )

  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length),
                 max = max(Sepal.Length)) %>%
      stat_summarise(.cols = c("ok" = "mean", "max"), stat = "sum", .by = mean),
    data.table::data.table(ok = 5822,
                           max = 7900)
  )

  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length),
                 max = max(Sepal.Length)) %>%
      stat_summarise(ok = mean, max, stat = "sum", .by = mean),
    data.table::data.table(mean = mean_sl,
                           ok = 5822,
                           max = 7900)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length),
                 max = max(Sepal.Length)) %>%
      stat_summarise(ok = mean, max, stat = "sum", .by = mean, .names = "{.fn}_{.col}"),
    data.table::data.table(mean = mean_sl,
                           sum_ok = 5822,
                           sum_max = 7900)
  )
  testthat::expect_equal(
    iris2 %>%
      transmute2(mean = mean(Sepal.Length),
                 max = max(Sepal.Length)) %>%
      stat_summarise(dplyr::across(c(ok = mean, max)), stat = "sum", .by = mean, .names = "{.fn}_{.col}"),
    data.table::data.table(mean = mean_sl,
                           sum_ok = 5822,
                           sum_max = 7900)
  )
})