tests/testthat/test-dgc.R

In ClustMC: Cluster-Based Multiple Comparisons

test_that("console is used properly", {
  expect_output(dgc_test(PlantGrowth$weight, PlantGrowth$group))
  expect_output(
    dgc_test(PlantGrowth$weight, PlantGrowth$group,
      console = FALSE
    ),
    regexp = NA
  )
})

test_that("wrong alpha value is detected", {
  expect_warning(dgc_test(PlantGrowth$weight, PlantGrowth$group, alpha = 999))
  expect_no_warning(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    alpha = 0.05
  ))
  expect_no_warning(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    alpha = 0.01
  ))
})

test_that("alerts are shown properly", {
  expect_no_message(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    alpha = 0.05
  ))
  expect_message(dgc_test(PlantGrowth$weight, PlantGrowth$group, alpha = 0.01),
    regexp = "No differences"
  )
  expect_message(dgc_test(iris$Sepal.Length, iris$Species),
    regexp = "All differences"
  )
})

test_that("error due to non-existent column is displayed correctly", {
  modelo <- lm(data = PlantGrowth, weight ~ group)
  expect_no_error(dgc_test(modelo, "group"))
  expect_error(dgc_test(modelo, "columna_inexistente"),
    regexp = "Column.*can't be found in"
  )
})

test_that("errors due to k and n are displayed correctly", {
  expect_error(dgc_test(mtcars$mpg, mtcars$vs),
    regexp = "at least.*treatments"
  )
  expect_error(dgc_test(mtcars$mpg, mtcars$wt),
    regexp = "at least.*observations"
  )
})

test_that("wrong object type or length returns error", {
  expect_error(dgc_test(PlantGrowth$weight[1:29], PlantGrowth$group),
    regexp = "variable lengths"
  )
  expect_error(dgc_test(PlantGrowth$weight, PlantGrowth$group[1:29]),
    regexp = "variable lengths"
  )
  expect_error(dgc_test(PlantGrowth$weight, "PlantGrowth$group"))
  expect_error(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    show_plot = "Yes"
  ))
  expect_error(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    console = "Yes"
  ))
  expect_error(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    abline_options = 2
  ))
})

test_that("plots are properly rendered", {
  expect_no_error(dgc_test(PlantGrowth$weight, PlantGrowth$group, col = "red"))
  expect_no_error(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    abline_options = list(col = "orange")
  ))
  expect_no_error(dgc_test(PlantGrowth$weight, PlantGrowth$group,
    abline_options = list(col = "purple"), col = "green"
  ))
})

test_that("returns are correct", {
  expect_type(dgc_test(PlantGrowth$weight, PlantGrowth$group)$stats,
    type = "list"
  )
  expect_type(dgc_test(PlantGrowth$weight, PlantGrowth$group)$groups,
    type = "list"
  )
  expect_type(dgc_test(PlantGrowth$weight, PlantGrowth$group)$groups[[1]],
    type = "integer"
  )
  expect_type(dgc_test(PlantGrowth$weight, PlantGrowth$group)$parameters,
    type = "list"
  )
  expect_length(dgc_test(
    PlantGrowth$weight,
    PlantGrowth$group
  )$parameters, n = 5)

  if (length(dgc_test(PlantGrowth$weight, PlantGrowth$group)$parameters) == 5) {
    expect_type(dgc_test(PlantGrowth$weight, PlantGrowth$group)$parameters[[1]],
      type = "integer"
    )
    for (i in 2:4) {
      expect_type(
        dgc_test(PlantGrowth$weight, PlantGrowth$group)$parameters[2:5][[i]],
        type = "double"
      )
    }
  }

  expect_s3_class(
    dgc_test(PlantGrowth$weight, PlantGrowth$group)$dendrogram_data,
    "hclust"
  )
})