tests/testthat/test-glm.R

In insight: Easy Access to Model Information for Various Model Objects

skip_if_not_installed("glmmTMB")

data(Salamanders, package = "glmmTMB")
Salamanders$cover <- abs(Salamanders$cover)
dat <<- Salamanders

m1 <- glm(count ~ mined + log(cover) + sample,
  family = poisson,
  data = dat
)

test_that("model_info", {
  expect_true(model_info(m1)$is_poisson)
  expect_true(model_info(m1)$is_count)
  expect_false(model_info(m1)$is_negbin)
  expect_false(model_info(m1)$is_binomial)
  expect_false(model_info(m1)$is_linear)
})

test_that("loglik", {
  expect_equal(get_loglikelihood(m1), logLik(m1), ignore_attr = TRUE)
})

test_that("get_df", {
  expect_equal(get_df(m1), df.residual(m1), ignore_attr = TRUE)
  expect_equal(get_df(m1, type = "model"), attr(logLik(m1), "df"), ignore_attr = TRUE)
})

test_that("get_df", {
  expect_equal(
    get_df(m1, type = "residual"),
    df.residual(m1),
    ignore_attr = TRUE
  )
  expect_equal(
    get_df(m1, type = "normal"),
    Inf,
    ignore_attr = TRUE
  )
  expect_equal(
    get_df(m1, type = "wald"),
    Inf,
    ignore_attr = TRUE
  )
})


test_that("find_predictors", {
  expect_identical(find_predictors(m1), list(conditional = c("mined", "cover", "sample")))
  expect_identical(
    find_predictors(m1, flatten = TRUE),
    c("mined", "cover", "sample")
  )
  expect_null(find_predictors(m1, effects = "random"))
})

test_that("find_random", {
  expect_null(find_random(m1))
})

test_that("get_random", {
  expect_warning(get_random(m1))
})

test_that("find_response", {
  expect_identical(find_response(m1), "count")
})

test_that("get_response", {
  expect_identical(get_response(m1), Salamanders$count)
})

test_that("get_predictors", {
  expect_identical(colnames(get_predictors(m1)), c("mined", "cover", "sample"))
})

test_that("link_inverse", {
  expect_equal(link_inverse(m1)(0.2), exp(0.2), tolerance = 1e-5)
})

test_that("linkfun", {
  expect_equal(link_function(m1)(0.2), -1.609438, tolerance = 1e-4)
})

test_that("get_data", {
  expect_identical(nrow(get_data(m1)), 644L)
  expect_identical(
    colnames(get_data(m1)),
    c("count", "mined", "cover", "sample")
  )
})

test_that("get_call", {
  expect_true(inherits(get_call(m1), "call")) # nolint
})

test_that("find_formula", {
  expect_length(find_formula(m1), 1)
  expect_equal(
    find_formula(m1),
    list(conditional = as.formula("count ~ mined + log(cover) + sample")),
    ignore_attr = TRUE
  )
})

test_that("find_variables", {
  expect_identical(
    find_variables(m1),
    list(
      response = "count",
      conditional = c("mined", "cover", "sample")
    )
  )
  expect_identical(
    find_variables(m1, flatten = TRUE),
    c("count", "mined", "cover", "sample")
  )
})

test_that("n_obs", {
  expect_identical(n_obs(m1), 644L)
})

test_that("find_parameters", {
  expect_identical(
    find_parameters(m1),
    list(
      conditional = c("(Intercept)", "minedno", "log(cover)", "sample")
    )
  )
  expect_identical(nrow(get_parameters(m1)), 4L)
  expect_identical(
    get_parameters(m1)$Parameter,
    c("(Intercept)", "minedno", "log(cover)", "sample")
  )
})

test_that("is_multivariate", {
  expect_false(is_multivariate(m1))
})

test_that("find_terms", {
  expect_identical(
    find_terms(m1),
    list(
      response = "count",
      conditional = c("mined", "log(cover)", "sample")
    )
  )
})

test_that("find_algorithm", {
  expect_identical(find_algorithm(m1), list(algorithm = "ML"))
})

test_that("find_statistic", {
  expect_identical(find_statistic(m1), "z-statistic")
})

test_that("get_statistic", {
  expect_equal(
    get_statistic(m1)$Statistic,
    c(
      -10.7066515607315,
      18.1533878215937,
      -1.68918157150882,
      2.23541768590273
    ),
    tolerance = 1e-4
  )
})

test_that("model_info, bernoulli", {
  skip_if_not_installed("lme4")
  data(cbpp, package = "lme4")
  m <- glm(
    cbind(incidence, size - incidence) ~ size + period,
    family = binomial(),
    data = cbpp
  )
  info <- model_info(m)
  expect_true(info$is_binomial)
  expect_false(info$is_bernoulli)
  expect_true(info$is_logit)
  expect_true(info$is_trial)
  expect_identical(info$family, "binomial")

  data(mtcars)
  m <- glm(
    am ~ cyl,
    family = binomial(),
    data = mtcars
  )
  info <- model_info(m)
  expect_true(info$is_binomial)
  expect_true(info$is_bernoulli)
  expect_true(info$is_logit)
  expect_false(info$is_trial)
  expect_identical(info$family, "binomial")
})