tests/testthat/test-lavaan.R

In parameters: Processing of Model Parameters

skip_if_not_installed("lavaan")

data(PoliticalDemocracy, package = "lavaan")

model <- "
    # measurement model
      ind60 =~ x1 + x2 + x3
      dem60 =~ y1 + y2 + y3 + y4
      dem65 =~ y5 + y6 + y7 + y8
    # regressions
      dem60 ~ ind60
      dem65 ~ ind60 + dem60
    # residual correlations
      y1 ~~ y5
      y2 ~~ y4 + y6
      y3 ~~ y7
      y4 ~~ y8
      y6 ~~ y8
  "

m <- lavaan::sem(model, data = PoliticalDemocracy, test = "Satorra-Bentler")

test_that("unstandardized", {
  mp <- model_parameters(m, eta_squared = "raw")
  ml <- lavaan::parameterEstimates(m, se = TRUE)
  ml <- ml[(ml$lhs != ml$rhs) & (ml$op != "~1"), ]

  expect_equal(mp$Coefficient, ml$est, tolerance = 1e-3)
  expect_equal(mp$SE, ml$se, tolerance = 1e-3)
})

test_that("standardized", {
  mp <- model_parameters(m, standardize = TRUE)
  ml <- lavaan::standardizedSolution(m, type = "std.all", se = TRUE)
  ml <- ml[(ml$lhs != ml$rhs) & (ml$op != "~1"), ]

  expect_equal(mp$Coefficient, ml$est, tolerance = 1e-3)
  expect_equal(mp$SE, ml$se, tolerance = 1e-3)
})

test_that("standardized-lv", {
  mp <- model_parameters(m, standardize = "latent")
  ml <- lavaan::standardizedSolution(m, type = "std.lv", se = TRUE)
  ml <- ml[(ml$lhs != ml$rhs) & (ml$op != "~1"), ]

  expect_equal(mp$Coefficient, ml$est, tolerance = 1e-3)
  expect_equal(mp$SE, ml$se, tolerance = 1e-3)
})

test_that("standardized-nox", {
  mp <- model_parameters(m, standardize = "no_exogenous")
  ml <- lavaan::standardizedSolution(m, type = "std.nox", se = TRUE)
  ml <- ml[(ml$lhs != ml$rhs) & (ml$op != "~1"), ]

  expect_equal(mp$Coefficient, ml$est, tolerance = 1e-3)
  expect_equal(mp$SE, ml$se, tolerance = 1e-3)
})

test_that("standardized no CI", {
  mod <- lavaan::cfa("ind60 =~ x1 + x2 + x3", data = PoliticalDemocracy)
  p <- parameters(mod, standardize = "all", ci = NULL)
  expect_s3_class(p, "parameters_sem")
})

test_that("simulate_model and equivalence_test work for lavaan", {
  skip_on_cran()
  skip_if_not_installed("bayestestR")
  skip_if_not_installed("lavaan")
  skip_if_not_installed("insight", minimum_version = "1.5.2")

  # Create simple example data
  set.seed(123)
  n <- 100
  x <- rnorm(n)
  y <- 0.3 * x + rnorm(n, 0, 0.8) # Small effect: β = 0.3
  data <- data.frame(x = x, y = y)

  # Fit simple lavaan model
  model <- 'y ~ x'
  m <- lavaan::sem(model, data = data)

  expect_s3_class(simulate_model(m, iterations = 10), "parameters_simulate_model")
  expect_s3_class(
    equivalence_test(m, iterations = 10, verbose = FALSE),
    "equivalence_test"
  )

  set.seed(123)
  out <- equivalence_test(m, range = c(-0.2, 0.2))
  expect_identical(
    capture.output(out),
    c(
      "# Test for Practical Equivalence",
      "",
      "  ROPE: [-0.20 0.20]",
      "",
      "Parameter |        H0 | inside ROPE |      95% HDI",
      "--------------------------------------------------",
      "y~x       | Undecided |     23.05 % | [0.09, 0.42]",
      "y~~y      |  Rejected |      0.00 % | [0.43, 0.76]",
      "",
      ""
    )
  )
})

test_that("equivalence_test finds a lavaan model defined inside a function", {
  skip_on_cran()
  skip_if_not_installed("bayestestR")
  skip_if_not_installed("lavaan")

  # the default ROPE range needs the model, which is looked up by name
  # in the calling frames
  fit_and_test <- function() {
    set.seed(123)
    d <- data.frame(x = rnorm(100))
    d$y <- 0.3 * d$x + rnorm(100, 0, 0.8)
    m <- lavaan::sem("y ~ x", data = d)
    equivalence_test(m, iterations = 10, verbose = FALSE)
  }
  out <- fit_and_test()
  expect_s3_class(out, "equivalence_test")
  expect_identical(out$Parameter, c("y~x", "y~~y"))
})