test-convert_between.R
In effectsize: Indices of Effect Size

# OR, d, r ------------------------------

test_that("oddsratio_to_d", {
  expect_equal(oddsratio_to_d(0.2), -0.887, tolerance = 0.01)
  expect_equal(oddsratio_to_d(-1.45, log = TRUE), -0.7994, tolerance = 0.01)
  expect_equal(d_to_oddsratio(-0.887), 0.2, tolerance = 0.01)
  expect_equal(d_to_oddsratio(-0.7994, log = TRUE), -1.45, tolerance = 0.01)
})

test_that("oddsratio_to_d (exact)", {
  d <- c(0, 0.2, 0.5, 0.8)
  p0 <- pnorm(0, lower.tail = FALSE)
  p1 <- pnorm(0, mean = d, lower.tail = FALSE)
  OR <- probs_to_odds(p1) / probs_to_odds(p0)

  expect_equal(cor(oddsratio_to_d(OR), d), 1, tolerance = 0.0001)
  expect_equal(oddsratio_to_d(1), 0, tolerance = 0.0001)
  expect_equal(oddsratio_to_d(OR, p0), d, tolerance = 0.0001)

  expect_equal(cor(oddsratio_to_r(OR), d_to_r(d)), 1, tolerance = 0.0002)
  expect_equal(oddsratio_to_r(1), 0, tolerance = 0.0001)
  expect_equal(oddsratio_to_r(OR, p0), d_to_r(d), tolerance = 0.0001)

  # From Chen et al 2010
  chen_tab_1 <- as.matrix(
    read.table(
      text = "p0    OR_1    OR_2    OR_3
        0.0100  1.6814  3.4739  6.7128
        0.0200  1.6146  3.1332  5.7486
        0.0300  1.5733  2.9535  5.2592
        0.0400  1.5455  2.8306  4.9471
        0.0500  1.5228  2.7416  4.7233
        0.0600  1.5060  2.6741  4.5536
        0.0700  1.4926  2.6177  4.4191
        0.0800  1.4811  2.5707  4.3097
        0.0900  1.4709  2.5309  4.2167
        0.1000  1.4615  2.4972  4.1387",
      header = TRUE
    )
  )

  for (i in seq_len(nrow(chen_tab_1))) {
    d_recovered <- oddsratio_to_d(chen_tab_1[i, 2:4], p0 = chen_tab_1[i, 1])
    expect_equal(
      d_recovered,
      c(0.2, 0.5, 0.8),
      tolerance = 0.01,
      ignore_attr = TRUE
    )
  }
})


test_that("d_to_r", {
  expect_equal(d_to_r(1.1547), 0.5, tolerance = 0.01)
  expect_equal(r_to_d(0.5), 1.1547, tolerance = 0.01)

  expect_equal(
    oddsratio_to_r(d_to_oddsratio(r_to_d(0.5))),
    0.5,
    tolerance = 0.001
  )
  expect_equal(
    oddsratio_to_d(r_to_oddsratio(d_to_r(1), log = TRUE), log = TRUE),
    1,
    tolerance = 0.001
  )

  r <- cor(mtcars$hp, -mtcars$am)
  d <- cohens_d(hp ~ am, data = mtcars, ci = NULL)[[1]]
  n <- table(mtcars$am)
  expect_gt(r_to_d(r), d)
  expect_lt(d_to_r(d), r)
  expect_equal(r_to_d(r, n[1], n[2]), d, ignore_attr = TRUE)
  expect_equal(d_to_r(d, n[1], n[2]), r, ignore_attr = TRUE)

  expect_identical(d_to_r(0.5, n1 = 10), d_to_r(0.5, 10, 10))
  expect_identical(d_to_r(0.5, n2 = 10), d_to_r(0.5, 10, 10))
})

test_that("OR and logOR <=> d, r", {
  expect_equal(
    oddsratio_to_d(3),
    logoddsratio_to_d(log(3)),
    tolerance = 1e-4
  )
  expect_equal(
    log(d_to_oddsratio(3)),
    d_to_logoddsratio(3),
    tolerance = 1e-4
  )
  expect_equal(
    oddsratio_to_r(2),
    logoddsratio_to_r(log(2)),
    tolerance = 1e-4
  )
  expect_equal(
    log(r_to_oddsratio(0.5)),
    r_to_logoddsratio(0.5),
    tolerance = 1e-4
  )
})

# OR, RR, ARR, NNT -------------------------

test_that("OR, RR, ARR, NNT | numeric", {
  p0 <- 0.4
  p1 <- 0.7

  OR <- probs_to_odds(p1) / probs_to_odds(p0)
  RR <- p1 / p0
  ARR <- p1 - p0
  NNT <- 1 / ARR

  expect_equal(nnt_to_arr(NNT), ARR, tolerance = 1e-4)
  expect_equal(arr_to_nnt(ARR), NNT, tolerance = 1e-4)

  expect_equal(riskratio_to_oddsratio(RR, p0 = p0), OR, tolerance = 1e-4)
  expect_equal(oddsratio_to_riskratio(OR, p0 = p0), RR, tolerance = 1e-4)
  expect_equal(
    oddsratio_to_riskratio(1 / OR, p0 = p1),
    1 / RR,
    tolerance = 1e-4
  )
  expect_equal(riskratio_to_arr(RR, p0 = p0), ARR, tolerance = 1e-4)
  expect_equal(oddsratio_to_arr(OR, p0 = p0), ARR, tolerance = 1e-4)
  expect_equal(arr_to_oddsratio(ARR, p0 = p0), OR, tolerance = 1e-4)
  expect_equal(arr_to_riskratio(ARR, p0 = p0), RR, tolerance = 1e-4)

  expect_equal(
    riskratio_to_oddsratio(RR, p0 = p0, log = TRUE),
    log(OR),
    tolerance = 1e-4
  )
  expect_equal(
    oddsratio_to_riskratio(log(OR), p0 = p0, log = TRUE),
    RR,
    tolerance = 1e-4
  )
  expect_equal(
    arr_to_oddsratio(ARR, p0 = p0, log = TRUE),
    log(OR),
    tolerance = 1e-4
  )
  expect_equal(
    oddsratio_to_arr(log(OR), p0 = p0, log = TRUE),
    ARR,
    tolerance = 1e-4
  )
})


test_that("OR <=> RR | models", {
  skip_on_cran()

  # -- GLMs --
  data(mtcars)

  m1 <<- glm(am ~ factor(cyl), data = mtcars, family = binomial("logit"))

  expect_warning(RR <- oddsratio_to_riskratio(m1, ci = NULL), "p0") # nolint
  expect_false("(Intercept)" %in% RR$Parameter)
  expect_true("(p0)" %in% RR$Parameter)

  expect_warning(
    RR <- oddsratio_to_riskratio(m1, ci_method = "wald", p0 = 0.7272727),
    NA
  ) # nolint
  expect_false("(Intercept)" %in% RR$Parameter)
  expect_true("(p0)" %in% RR$Parameter)
  # these values confirmed from emmeans
  expect_equal(RR$Coefficient, c(0.7272, 0.5892, 0.1964), tolerance = 0.001)
  expect_equal(RR$CI_low, c(NA, 0.1267, 0.0303), tolerance = 0.001)
  expect_equal(RR$CI_high, c(NA, 1.1648, 0.7589), tolerance = 0.001)

  expect_warning(RR <- oddsratio_to_riskratio(m1, p0 = 0.05), NA) # nolint
  expect_true("(p0)" %in% RR$Parameter)
  expect_false("(Intercept)" %in% RR$Parameter)
  # these values confirmed from emmeans
  expect_equal(RR$Coefficient, c(0.05, 0.29173, 0.06557), tolerance = 0.001)

  # -- GLMs2 --
  data(mtcars)

  m2 <<- glm(am ~ factor(cyl), data = mtcars, family = binomial("log"))

  ORt <- parameters::model_parameters(m1, exp = TRUE)

  expect_warning(OR <- riskratio_to_oddsratio(m2, ci = NULL), "p0") # nolint
  expect_false("(Intercept)" %in% OR$Parameter)
  expect_true("(p0)" %in% OR$Parameter)

  expect_warning(
    OR <- riskratio_to_oddsratio(m2, ci_method = "wald", p0 = 0.7272727),
    NA
  ) # nolint
  expect_false("(Intercept)" %in% OR$Parameter)
  expect_true("(p0)" %in% OR$Parameter)
  # these values confirmed from marginaleffects
  expect_equal(OR$Coefficient[-1], ORt$Coefficient[-1], tolerance = 0.001)
  expect_equal(OR$CI_low, c(NA, 0.07642535, 0.01467661), tolerance = 0.001)
  expect_equal(OR$CI_high, c(NA, NA, 0.443517), tolerance = 0.001)

  expect_warning(OR <- riskratio_to_oddsratio(m2, p0 = 0.05), NA) # nolint
  expect_true("(p0)" %in% OR$Parameter)
  expect_false("(Intercept)" %in% OR$Parameter)
  # these values confirmed from marginaleffects
  expect_equal(OR$Coefficient, c(0.05, 0.5768169, 0.1884581), tolerance = 0.001)

  # -- GLMMs --
  skip_if_not_installed("lme4")
  m <<- lme4::glmer(
    am ~ factor(cyl) + (1 | gear),
    data = mtcars,
    family = binomial()
  )

  expect_warning(RR <- oddsratio_to_riskratio(m, ci = NULL), "p0") # nolint
  expect_false("(Intercept)" %in% RR$Parameter)
  expect_true("(p0)" %in% RR$Parameter)

  expect_warning(
    RR <- oddsratio_to_riskratio(m, ci_method = "wald", p0 = 0.7047536),
    NA
  ) # nolint
  expect_false("(Intercept)" %in% RR$Parameter)
  expect_true("(p0)" %in% RR$Parameter)
  # these values confirmed from emmeans
  expect_equal(RR$Coefficient, c(0.7048, 0.6042, 0.4475), tolerance = 0.001)
  expect_equal(RR$CI_low, c(NA, 0.08556, 0.0102), tolerance = 0.001)
  expect_equal(RR$CI_high, c(NA, 1.2706, 1.3718), tolerance = 0.001)
})

test_that("OR and logOR <=> ARR", {
  expect_equal(
    log(arr_to_oddsratio(0.2, p0 = 0.3)),
    arr_to_logoddsratio(0.2, p0 = 0.3),
    tolerance = 1e-4
  )
  expect_equal(
    oddsratio_to_arr(2, p0 = 0.3),
    logoddsratio_to_arr(log(2), p0 = 0.3),
    tolerance = 1e-4
  )
})

test_that("=> probs", {
  p0 <- 0.4
  p1 <- 0.7

  OR <- probs_to_odds(p1) / probs_to_odds(p0)
  RR <- p1 / p0
  ARR <- p1 - p0
  NNT <- arr_to_nnt(ARR)

  expect_equal(riskratio_to_probs(RR, p0 = p0), p1)
  expect_equal(oddsratio_to_probs(OR, p0 = p0), p1)

  expect_equal(nnt_to_probs(NNT, p0 = p0, odds = TRUE), probs_to_odds(p1))

  expect_equal(arr_to_probs(-ARR, p0 = p1), p0)
  expect_equal(nnt_to_probs(-NNT, p0 = p1), p0)
})

# Odds, p -------------------------------

test_that("odds_to_probs", {
  expect_equal(odds_to_probs(3), 0.75, tolerance = 0.01)
  expect_equal(probs_to_odds(0.75), 3, tolerance = 0.01)
  expect_equal(probs_to_odds(0.75, log = TRUE), 1.098, tolerance = 0.01)
  expect_equal(odds_to_probs(1.098, log = TRUE), 0.75, tolerance = 0.01)

  # Data frames
  expect_error(odds_to_probs(iris), "deprecated")
})

# ANOVA ------------------------------

test_that("between anova", {
  expect_equal(eta2_to_f2(0.25), 1 / 3, tolerance = 1e-4)
  expect_equal(eta2_to_f(0.25), sqrt(eta2_to_f2(0.25)), tolerance = 1e-4)

  expect_equal(f2_to_eta2(1 / 3), 0.25)
  expect_equal(f_to_eta2(1 / sqrt(3)), f2_to_eta2(1 / 3), tolerance = 1e-4)
})