tests/testthat/test-multz.R

In ssutil: Sample Size Calculation Tools

# Define your expectations from multz function
# Expected values from table B.1 and B.2 from  Bechhofer, Design and analysis of
# experiments for statistical selection, screening and multiple comparisons book

expect_multz <- function(alpha, n, rho, expected_result, tol){
  output <- multz(alpha, n, rho, lower.tail = FALSE)

  # Compare the result with an expected value by using the expect_equal function
  expect_equal(output, expected_result, tolerance = tol)
}

test_that("multz function calculates correct z-values", {
  expect_multz(0.01, 1, 1, 2.326, 0.001)
  expect_multz(0.1, 3, 0.5, 1.734, 0.001)
  expect_multz(0.2, 7, 0.5, 1.655, 0.001)
  expect_multz(0.1, 7, 0.1, 2.16, 0.01)
  expect_multz(0.01, 24, 0.1, 3.34, 0.01)
  expect_multz(0.1, 7, 0.3, 2.11, 0.01)
  expect_multz(0.01, 24, 0.3, 3.32, 0.01)
  expect_multz(0.1, 7, 0.7, 1.90, 0.01)
  expect_multz(0.01, 24, 0.7, 3.13, 0.01)
})


multz_eq_multp <- function(alpha, n, rho, tol){
  z <- multz(alpha, n, rho, lower.tail = FALSE)
  expect_equal(multp(z, n, rho, lower.tail = FALSE), alpha, tolerance = tol)
}

test_that("multz and multp agrees",{
  multz_eq_multp(0.01, 1, 1,  0.01)
  multz_eq_multp(0.1, 3, 0.5, 0.01)
  multz_eq_multp(0.2, 7, 0.5, 0.01)
  multz_eq_multp(0.1, 7, 0.1, 0.01)
  multz_eq_multp(0.01, 24, 0.1, 0.01)
  multz_eq_multp(0.1, 7, 0.3, 0.01)
  multz_eq_multp(0.01, 24, 0.3, 0.01)
  multz_eq_multp(0.1, 7, 0.7, 0.01)
  multz_eq_multp(0.01, 24, 0.7, 0.01)
})