tests/testthat/test-cor_multilevel.R

In easystats/correlation: Methods for Correlation Analysis

test_that("comparison rmcorr", {
  skip_if_not_or_load_if_installed("lme4")
  skip_if_not_or_load_if_installed("rmcorr")
  set.seed(123)
  rez_rmcorr <- rmcorr::rmcorr(Species, Sepal.Length, Sepal.Width, dataset = iris)

  set.seed(123)
  rez <- cor_test(iris[c(1, 2, 5)], "Sepal.Length", "Sepal.Width", partial = TRUE, multilevel = TRUE)

  expect_equal(rez$r, rez_rmcorr$r, tolerance = 0.001)
  expect_equal(rez$p, rez_rmcorr$p, tolerance = 0.001)
  # expect_equal(rez$df_error, rez_rmcorr$df)
  expect_equal(rez$CI_low, rez_rmcorr$CI[1], tolerance = 0.01)
  expect_equal(rez$CI_high, rez_rmcorr$CI[2], tolerance = 0.01)
})


test_that("Reductio ad absurdum", {
  skip_if_not_or_load_if_installed("lme4")
  cormatrix <- matrix(
    c(
      1.0, 0.3, 0.6,
      0.3, 1.0, 0.0,
      0.6, 0.0, 1.0
    ),
    nrow = 3
  )

  data <- bayestestR::simulate_correlation(n = 500, r = cormatrix)
  # Add factor levels "at random", so the grouping structure should NOT change much
  data$Group <- sample(rep_len(c("A", "B", "C"), length.out = 500))

  rez <- correlation(data)
  expect_equal(max(as.matrix(rez) - cormatrix), 0, tolerance = 0.000001)

  rez <- suppressMessages(correlation(data, multilevel = TRUE, verbose = FALSE))
  expect_equal(max(as.matrix(rez) - cormatrix), 0, tolerance = 0.01)

  rez <- suppressMessages(correlation(data, multilevel = TRUE, partial = TRUE, verbose = FALSE))
  expect_equal(max(as.matrix(pcor_to_cor(rez)) - cormatrix), 0, tolerance = 0.01)
})