tests/testthat/test-print_test_predictions.R

In strengejacke/ggeffects: Create Tidy Data Frames of Marginal Effects for 'ggplot' from Model Outputs

skip_on_os(c("mac", "linux", "solaris"))
skip_if_not_installed("marginaleffects")
skip_if_not_installed("ggplot2")
skip_if_not_installed("lme4")

set.seed(123)
dat <- data.frame(
  outcome = rbinom(n = 100, size = 1, prob = 0.35),
  var_binom = as.factor(rbinom(n = 100, size = 1, prob = 0.2)),
  var_cont = rnorm(n = 100, mean = 10, sd = 7),
  groups = sample(letters[1:4], size = 100, replace = TRUE)
)

m <- glm(outcome ~ var_binom * var_cont + groups,
  data = dat, family = binomial()
)

test_that("print hypothesis_test simple contrast link scale", {
  out <- hypothesis_test(m, "var_binom", scale = "link")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test simple predictions link scale", {
  out <- hypothesis_test(m, "var_binom", test = NULL, scale = "link")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test simple contrast exp scale", {
  out <- hypothesis_test(m, "var_binom", scale = "exp")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test simple contrast odds ratio scale", {
  out <- hypothesis_test(m, "var_binom", scale = "oddsratios")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test simple contrast response scale", {
  out <- hypothesis_test(m, "var_binom", scale = "response")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test simple predictions exp scale", {
  out <- hypothesis_test(m, "var_binom", test = NULL, scale = "exp")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test simple predictions odds ratio scale", {
  out <- hypothesis_test(m, "var_binom", test = NULL, scale = "oddsratios")
  expect_snapshot(print(out))
})

test_that("print hypothesis_test contrasts link scale", {
  out <- hypothesis_test(m, c("var_binom", "var_cont"), scale = "link")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test predictions link scale", {
  out <- hypothesis_test(m, c("var_binom", "var_cont"), test = NULL, scale = "link")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test contrasts exp scale", {
  out <- hypothesis_test(m, c("var_binom", "var_cont"), scale = "exp")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test contrasts response scale", {
  out <- hypothesis_test(m, c("var_binom", "var_cont"), scale = "response")
  expect_snapshot(print(out))
})
test_that("print hypothesis_test predictions exp scale", {
  out <- hypothesis_test(m, c("var_binom", "var_cont"), test = NULL, scale = "exp")
  expect_snapshot(print(out))
})

test_that("print hypothesis_test many rows", {
  dat <- mtcars
  dat$gear <- factor(dat$gear)
  dat$vs <- factor(dat$vs)
  dat$cyl <- factor(dat$cyl)
  mod <- lm(mpg ~ cyl * vs * gear, data = dat)
  ht <- suppressWarnings(hypothesis_test(
    mod,
    terms = c("cyl", "vs", "gear"),
    test = "(b1 - b13) = (b3 - b15)"
  ))
  out <- capture.output(print(ht))
  expect_identical(
    out,
    c(
      "Hypothesis        | Contrast |       95% CI |     p",
      "---------------------------------------------------",
      "(b1-b13)=(b3-b15) |    -8.55 | -19.86, 2.76 | 0.131",
      "",
      "Tested hypothesis: (cyl[4],vs[0],gear[3] - cyl[4],vs[0],gear[5]) =",
      "  (cyl[8],vs[0],gear[3] - cyl[8],vs[0],gear[5])"
    )
  )
  expect_identical(
    attributes(ht)$hypothesis_label,
    "(cyl[4],vs[0],gear[3] - cyl[4],vs[0],gear[5]) = (cyl[8],vs[0],gear[3] - cyl[8],vs[0],gear[5])"
  )
  # check that operators are not replaced if inside brackets
  dat <- mtcars
  dat$gear <- factor(dat$gear)
  dat$vs <- factor(dat$vs)
  dat$cyl <- factor(dat$cyl)
  levels(dat$gear) <- c("-40", "41-64", "65+")
  levels(dat$vs) <- c("a=1", "b=2")
  mod <- lm(mpg ~ cyl * vs * gear, data = dat)
  ht <- suppressWarnings(hypothesis_test(
    mod,
    terms = c("cyl", "vs", "gear"),
    test = "(b1 - b13) = (b3 - b15)"
  ))
  out <- capture.output(print(ht))
  expect_identical(
    out,
    c(
      "Hypothesis        | Contrast |       95% CI |     p",
      "---------------------------------------------------",
      "(b1-b13)=(b3-b15) |    -8.55 | -19.86, 2.76 | 0.131",
      "",
      "Tested hypothesis: (cyl[4],vs[a=1],gear[-40] - cyl[4],vs[a=1],gear[65+])",
      "  = (cyl[8],vs[a=1],gear[-40] - cyl[8],vs[a=1],gear[65+])"
    )
  )
  expect_identical(
    attributes(ht)$hypothesis_label,
    "(cyl[4],vs[a=1],gear[-40] - cyl[4],vs[a=1],gear[65+]) = (cyl[8],vs[a=1],gear[-40] - cyl[8],vs[a=1],gear[65+])"
  )
  # check that collapse_levels works
  ht1 <- suppressWarnings(hypothesis_test(
    mod,
    terms = c("cyl", "vs", "gear"),
    by = "gear",
    collapse_levels = TRUE
  ))
  ht2 <- suppressWarnings(hypothesis_test(
    mod,
    terms = c("cyl", "vs", "gear"),
    by = "gear",
    collapse_levels = FALSE
  ))
  expect_equal(ht1$Contrast, ht2$Contrast, tolerance = 1e-3)
  expect_identical(
    ht1$vs,
    c(
      "a=1", "a=1", "a=1-b=2", "a=1-b=2", "a=1-b=2", "a=1", "a=1-b=2",
      "a=1-b=2", "a=1-b=2", "a=1-b=2", "a=1-b=2", "a=1-b=2", "b=2",
      "b=2", "b=2", "a=1", "a=1", "a=1-b=2", "a=1-b=2", "a=1-b=2",
      "a=1", "a=1-b=2", "a=1-b=2", "a=1-b=2", "a=1-b=2", "a=1-b=2",
      "a=1-b=2", "b=2", "b=2", "b=2", "a=1", "a=1", "a=1-b=2", "a=1-b=2",
      "a=1-b=2", "a=1", "a=1-b=2", "a=1-b=2", "a=1-b=2", "a=1-b=2",
      "a=1-b=2", "a=1-b=2", "b=2", "b=2", "b=2"
    )
  )
  expect_identical(
    ht1$cyl,
    c(
      "4-6", "4-8", "4", "4-6", "4-8", "6-8", "6-4", "6", "6-8",
      "8-4", "8-6", "8", "4-6", "4-8", "6-8", "4-6", "4-8", "4", "4-6",
      "4-8", "6-8", "6-4", "6", "6-8", "8-4", "8-6", "8", "4-6", "4-8",
      "6-8", "4-6", "4-8", "4", "4-6", "4-8", "6-8", "6-4", "6", "6-8",
      "8-4", "8-6", "8", "4-6", "4-8", "6-8"
    )
  )
})

test_that("print hypothesis_test comma and dash levels", {
  data(iris)
  d <- iris
  set.seed(123)
  d$f1 <- as.factor(sample(c("no comma", "with, comma", "and, another, comma"), nrow(d), replace = TRUE))
  d$f2 <- as.factor(sample(letters[1:2], nrow(d), replace = TRUE))

  m <- lme4::lmer(Sepal.Length ~ Sepal.Width + f1 + f2 + (1 | Species), data = d)
  ht <- hypothesis_test(m, c("f1", "f2"))
  expect_identical(nrow(ht), 15L)
  expect_snapshot(print(ht))

  d <- iris
  set.seed(123)
  d$f1 <- as.factor(sample(c("no comma", "with, comma", "and, another, comma"), nrow(d), replace = TRUE))
  d$f2 <- as.factor(sample(c("comma, here", "nothere"), nrow(d), replace = TRUE))

  m <- lme4::lmer(Sepal.Length ~ Sepal.Width + f1 + f2 + (1 | Species), data = d)
  ht <- hypothesis_test(m, c("f1", "f2"))
  expect_identical(nrow(ht), 15L)
  expect_snapshot(print(ht))

  d <- iris
  set.seed(1234)
  d$f1 <- as.factor(sample(c("no comma", "with, comma", "and, another, comma"), nrow(d), replace = TRUE))
  set.seed(123)
  d$f2 <- as.factor(sample(letters[1:2], nrow(d), replace = TRUE))

  m <- lme4::lmer(Sepal.Length ~ Sepal.Width + f1 + f2 + (1 | Species), data = d)
  ht <- hypothesis_test(m, c("Sepal.Width", "f1", "f2"))
  expect_identical(
    ht$f1,
    c(
      "and, another, comma-and, another, comma", "and, another, comma-no comma",
      "and, another, comma-no comma", "and, another, comma-with, comma",
      "and, another, comma-with, comma", "and, another, comma-no comma",
      "and, another, comma-no comma", "and, another, comma-with, comma",
      "and, another, comma-with, comma", "no comma-no comma", "no comma-with, comma",
      "no comma-with, comma", "no comma-with, comma", "no comma-with, comma",
      "with, comma-with, comma"
    )
  )
  expect_identical(
    ht$f2,
    c(
      "a-b", "a-a", "a-b", "a-a", "a-b", "b-a", "b-b", "b-a", "b-b",
      "a-b", "a-a", "a-b", "b-a", "b-b", "a-b"
    )
  )

  d <- iris
  set.seed(123)
  d$f1 <- as.factor(sample(c("no dash", "with, comma", "and-dash"), nrow(d), replace = TRUE))
  d$f2 <- as.factor(sample(c("comma, here", "dash-there"), nrow(d), replace = TRUE))

  m <- lme4::lmer(Sepal.Length ~ Sepal.Width + f1 + f2 + (1 | Species), data = d)
  ht <- hypothesis_test(m, c("f1", "f2"), collapse_levels = TRUE)
  expect_identical(nrow(ht), 15L)
  expect_snapshot(print(ht))
})

test_that("print hypothesis_test collapse levels", {
  data(efc, package = "ggeffects")
  efc$c172code <- as.factor(efc$c172code)
  efc$c161sex <- as.factor(efc$c161sex)
  levels(efc$c161sex) <- c("male", "female")
  m <- lm(barthtot ~ c12hour + neg_c_7 + c161sex * c172code, data = efc)

  out <- hypothesis_test(m, c("c172code", "c161sex"), collapse_levels = TRUE)
  expect_snapshot(print(out))
})

test_that("print hypothesis_test collapse CI", {
  data(efc, package = "ggeffects")
  efc$e42dep <- as.factor(efc$e42dep)
  fit <- lm(barthtot ~ e42dep + c160age, data = efc)
  pr <- ggpredict(fit, "e42dep")
  out <- hypothesis_test(pr)
  expect_snapshot(print(out))
  expect_snapshot(print(out, collapse_ci = TRUE))
})


test_that("hypothesis_test, ci-level", {
  data(iris)
  m <- lm(Sepal.Length ~ Species, data = iris)
  out <- hypothesis_test(m, "Species")
  expect_snapshot(print(out))
  out <- hypothesis_test(m, "Species", ci_level = 0.8)
  expect_snapshot(print(out))
})


test_that("glmmTMB, orderedbeta", {
  skip_if_not_installed("datawizard")
  skip_if_not_installed("glmmTMB")
  data(mtcars)
  mtcars$ord <- datawizard::normalize(mtcars$mpg)
  m <- glmmTMB::glmmTMB(
    ord ~ wt + hp + as.factor(gear) + (1 | cyl),
    data = mtcars,
    family = glmmTMB::ordbeta()
  )
  out2 <- predict_response(m, "gear", margin = "ame")
  expect_snapshot(print(test_predictions(out2)))
})