tests/testthat/test_bootstrap.R

In JonasMoss/univariateML: Maximum Likelihood Estimation for Univariate Densities

context("bootstrap")

# Gamma data.
set.seed(313)
x <- rgamma(10, 2, 3)
obj <- mlgamma(x)
obj2 <- mlexp(x)

# Different reducer and map.
set.seed(313)
expect_equal(
  c(bootstrapml(obj,
    reps = 2,
    map = function(x) x[1] / x[2],
    reducer = mean
  )),
  0.4352018,
  tolerance = 1e-5
)

# Different reducer and map.
set.seed(313)
expect_equal(
  object = unname(c(bootstrapml(obj, reps = 2, reducer = mean))),
  expected = c(4.333538, 10.581035),
  tolerance = 1e-5
)

# Common reducer
set.seed(313)
expect_equal(
  object = unname(c(bootstrapml(obj, reps = 2))),
  expected = c(2.369753, 5.001551, 6.297323, 16.160518),
  tolerance = 1e-5
)

# Keeping Names
set.seed(313)
expect_equal(
  object = colnames(bootstrapml(obj, reps = 2, reducer = mean)),
  expected = c("shape", "rate")
)

# Common reducer, one-dim
set.seed(313)
expect_equal(
  object = unname(c(bootstrapml(obj2, reps = 2))),
  expected = c(1.967853, 3.108297),
  tolerance = 1e-5
)