tests/testthat/test_mlcauchy.R

In univariateML: Maximum Likelihood Estimation for Univariate Densities

context("mlcauchy")

## Data generation.
set.seed(313)
tiny_data <- stats::rcauchy(10, 1, 7)
small_data <- stats::rcauchy(100, 10, 3)
medium_data <- stats::rcauchy(1000, 1 / 2, 2)
large_data <- stats::rcauchy(10000, 20, 13)

## Checks if the ML is correct.

mle1 <- suppressWarnings(nlm(function(p) {
  -sum(stats::dcauchy(tiny_data, p[1], p[2], log = TRUE))
}, p = c(1, 1)))

mle2 <- suppressWarnings(nlm(function(p) {
  -sum(stats::dcauchy(small_data, p[1], p[2], log = TRUE))
}, p = c(1, 1)))

mle3 <- suppressWarnings(nlm(function(p) {
  -sum(stats::dcauchy(medium_data, p[1], p[2], log = TRUE))
}, p = c(1, 1)))

mle4 <- suppressWarnings(nlm(function(p) {
  -sum(stats::dcauchy(large_data, p[1], p[2], log = TRUE))
}, p = c(
  stats::median(large_data),
  stats::median(abs(large_data - stats::median(large_data)))
)))

## Checks estimates.
expect_equal(mle1$estimate, as.numeric(mlcauchy(tiny_data)),
  tolerance = 1e-5
)
expect_equal(mle2$estimate, as.numeric(mlcauchy(small_data)),
  tolerance = 1e-5
)
expect_equal(mle3$estimate, as.numeric(mlcauchy(medium_data)),
  tolerance = 1e-5
)
expect_equal(mle4$estimate, as.numeric(mlcauchy(large_data)),
  tolerance = 1e-5
)

## Checks logLiks.
expect_equal(-mle1$minimum, attr(mlcauchy(tiny_data), "logLik"),
  tolerance = 1e-5
)
expect_equal(-mle2$minimum, attr(mlcauchy(small_data), "logLik"),
  tolerance = 1e-5
)
expect_equal(-mle3$minimum, attr(mlcauchy(medium_data), "logLik"),
  tolerance = 1e-5
)
expect_equal(-mle4$minimum, attr(mlcauchy(large_data), "logLik"),
  tolerance = 1e-5
)

## Finds errors with na and data out of bounds.
expect_error(mlcauchy(c(tiny_data, NA)))

## Checks that na.rm works as intended.
expect_equal(
  coef(mlcauchy(small_data)),
  coef(mlcauchy(c(small_data, NA), na.rm = TRUE))
)

## Check class.
est <- mlcauchy(small_data, na.rm = TRUE)
expect_equal(attr(est, "model"), "Cauchy")
expect_equal(class(est), "univariateML")