tests/testthat/test-integrated-simple.R

In clustTMB: Spatio-Temporal Finite Mixture Model using 'TMB'

library(clustTMB)
library(mclust)
library(sp)
library(testthat)
library(magrittr)
library(tidyr)
library(dplyr)
library(GGally)
# load data
library(palmerpenguins)

context("integrated simple test")

pen <- penguins_raw %>%
  dplyr::mutate(
    id = row_number(),
    `Culmen Length (mm)` = scale(`Culmen Length (mm)`),
    `Culmen Depth (mm)` = scale(`Culmen Depth (mm)`),
    `Flipper Length (mm)` = scale(`Flipper Length (mm)`),
    `Body Mass (g)` = scale(`Body Mass (g)`)
  ) %>%
  dplyr::select(
    id, Species, Region, Island, `Culmen Length (mm)`,
    `Culmen Depth (mm)`, `Flipper Length (mm)`,
    `Body Mass (g)`
  ) %>%
  tidyr::drop_na(`Culmen Length (mm)`)

dat <- pen %>%
  dplyr::select(
    `Culmen Length (mm)`, `Culmen Depth (mm)`,
    `Flipper Length (mm)`, `Body Mass (g)`
  ) %>%
  as.data.frame() %>%
  as.matrix()

mclust.mod <- Mclust(dat, G = 3, modelNames = "VII")

clustTMB.mod <- clustTMB(
  response = dat, G = 3,
  covariance.structure = "VII"
)


test_that("simple integrated test", {
  expect_equal(-mclust.mod$loglik,
    clustTMB.mod$opt$objective,
    tolerance = .01
  )
  expect_equal(1, MixSim::ClassProp(
    mclust.mod$classification,
    clustTMB.mod$report$classification
  ),
  tolerance = 0.01
  )
  expect_equal(1, MixSim::ClassProp(
    mclust.mod$classification,
    pen$Species
  ), tolerance = 0.1)
  expect_equal(1, MixSim::ClassProp(
    clustTMB.mod$report$classification,
    pen$Species
  ), tolerance = 0.1)
})

data("faithful")
mclust.mod <- Mclust(faithful, G = 2, modelNames = "VVV")
clustTMB.mod <- clustTMB(
  response = as.matrix(faithful),
  G = 2, covariance.structure = "VVV"
)

test_that("simple integrated test", {
  expect_equal(-mclust.mod$loglik,
    clustTMB.mod$opt$objective,
    tolerance = .0001
  )
  expect_equal(1, MixSim::ClassProp(
    mclust.mod$classification,
    clustTMB.mod$report$classification
  ))
})

# Add more tests for different clustTMB::fixStruct.names() using above two data sets
# can use first column of faithful to test "E" and "V"