tests/testthat/test-resampling_core.R

In BioTIMEr: Tools to Use and Explore the 'BioTIME' Database

# Create example data for testing resampling_core
set.seed(42)
test_dt <- data.table::data.table(
  assemblageID = rep(1:4, each = 192),
  minsamp = 2L,
  YEAR = rep(rep(2010:2015, each = 4), times = 8),
  SAMPLE_DESC = rep(LETTERS[1L:8L], each = 24),
  Species = unlist(lapply(
    X = 1L:8L,
    function(x) letters[sample(length(letters), 24L, replace = FALSE)]
  )),
  ABUNDANCE = rpois(24 * 8, 10),
  BIOMASS = abs(rnorm(24 * 8, mean = 0, sd = 1000))
)

resamps <- 3L

# Summarise = TRUE (default) ----
test_that("resampling_core returns a data frame", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = TRUE
  )
  expect_s3_class(result, "data.table")
})

test_that("resampling_core returns the expected column names - summarise = TRUE", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = TRUE
  )
  checkmate::expect_subset(
    x = colnames(result),
    choices = c(
      "assemblageID",
      "YEAR",
      "Species",
      "BIOMASS",
      "resamp"
    )
  )
})

test_that("resampling_core returns non-empty data frame", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = TRUE
  )
  expect_false(anyNA(result))
})

test_that("resampling_core returns consistent 'YEAR' values", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = TRUE
  )
  expected_years <- unique(test_dt$YEAR)
  expect_setequal(unique(result$YEAR), expected_years)
})

test_that("resampling_core returns consistent 'Species' values", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = TRUE
  )
  expected_species <- unique(test_dt$Species)
  expect_setequal(unique(result$Species), expected_species)
})


test_that("resampling_core returns positive 'BIOMASS' values", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = TRUE
  )
  checkmate::expect_numeric(
    result$BIOMASS,
    lower = 0L,
    null.ok = FALSE,
    all.missing = FALSE
  )
})

test_that("resampling_core summarise = TRUE returns consistent results", {
  skip_on_cran()
  skip_on_ci()

  expect_snapshot({
    set.seed(42)
    resampling_core(
      x = test_dt,
      measure = c("ABUNDANCE", "BIOMASS"),
      summarise = TRUE
    )
  })
})


# Summarise = FALSE ----
test_that("resampling_core summarise = FALSE returns a data frame", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = FALSE
  )
  expect_s3_class(result, "data.frame")
})

test_that("resampling_core returns the expected column names - summarise FALSE", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = FALSE
  )
  checkmate::expect_subset(
    x = colnames(result),
    choices = c(
      "YEAR",
      "Species",
      "BIOMASS",
      "SAMPLE_DESC",
      "assemblageID",
      "minsamp"
    )
  )
})

test_that("resampling_core returns non-empty data frame", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = FALSE
  )
  expect_false(anyNA(result))
})

test_that("resampling_core returns consistent 'YEAR' values", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = FALSE
  )
  expected_years <- unique(test_dt$YEAR)
  expect_setequal(unique(result$YEAR), expected_years)
})

test_that("resampling_core returns consistent 'Species' values", {
  result <- resampling_core(
    x = test_dt,
    measure = "BIOMASS",
    summarise = FALSE
  )
  expect_setequal(unique(result$Species), unique(test_dt$Species))
})


test_that("resampling_core summarise = FALSE returns consistent results", {
  skip_on_cran()
  skip_on_ci()

  expect_snapshot({
    set.seed(42)
    resampling_core(
      x = test_dt,
      measure = c("ABUNDANCE", "BIOMASS"),
      summarise = FALSE
    ) |>
      dplyr::as_tibble()
  })
})