tests/testthat/test-mplnDataGenerator.R

In anjalisilva/MPLNClust: Mixtures of Multivariate Poisson-Log Normal Model for Clustering Count Data

context("Checking for data simulation")
library(MPLNClust)

test_that("Data generation is as expected", {

  # Generating simulated data
  trueMu1 <- c(6.5, 6, 6, 6, 6, 6)
  trueMu2 <- c(2, 2.5, 2, 2, 2, 2)

  trueSigma1 <- diag(6) * 2
  trueSigma2 <- diag(6)

  set.seed(1234)
  simulatedCounts <- mplnDataGenerator(nObservations = 50,
                                        dimensionality = 6,
                                        mixingProportions = c(0.79, 0.21),
                                        mu = rbind(trueMu1, trueMu2),
                                        sigma = rbind(trueSigma1, trueSigma2),
                                        produceImage = "No")

  expect_type(simulatedCounts, "list")
  expect_length(simulatedCounts, 9)
  expect_s3_class(simulatedCounts, "mplnDataGenerator")
  expect_identical(trunc(simulatedCounts$observations), 50)
  expect_identical(trunc(simulatedCounts$dimensionality), 6)
  expect_named(simulatedCounts, c("dataset", "trueMembership",
                                  "probaPost", "trueNormFactors",
                                  "observations", "dimensionality",
                                  "mixingProportions", "trueMu",
                                  "trueSigma"))
  expect_output(str(simulatedCounts), "List of 9")
  expect_vector(simulatedCounts$mixingProportions, ptype = double(), size = 2)
})


context("Checking for invalid user input")
test_that("Data generate error upon invalid user input", {

  # nObservations provided as character
  expect_error(mplnDataGenerator(nObservations = "50",
                                        dimensionality = 6,
                                        mixingProportions = c(0.79, 0.21),
                                        mu = rbind(trueMu1, trueMu2),
                                        sigma = rbind(trueSigma1, trueSigma2),
                                        produceImage = "No"))


  # Generating simulated data - mu has incorrect dimension
  trueMu1 <- c(6.5, 6, 6, 6, 6)
  trueMu2 <- c(2, 2.5, 2, 2, 2)

  expect_error(mplnDataGenerator(nObservations = 50,
    dimensionality = 6,
    mixingProportions = c(0.79, 0.21),
    mu = rbind(trueMu1, trueMu2),
    sigma = rbind(trueSigma1, trueSigma2),
    produceImage = "No"))


  # Generating simulated data - sigma has incorrect dimension
  trueMu1 <- c(6.5, 6, 6, 6, 6, 6)
  trueMu2 <- c(2, 2.5, 2, 2, 2, 2)
  trueSigma1 <- diag(5) * 2
  trueSigma2 <- diag(5)

  expect_error(mplnDataGenerator(nObservations = 50,
    dimensionality = 6,
    mixingProportions = c(0.79, 0.21),
    mu = rbind(trueMu1, trueMu2),
    sigma = rbind(trueSigma1, trueSigma2),
    produceImage = "No"))


  # mixingProportions does not sum to 1
  expect_error(mplnDataGenerator(nObservations = 50,
    dimensionality = 6,
    mixingProportions = c(0.79, 0.2),
    mu = rbind(trueMu1, trueMu2),
    sigma = rbind(trueSigma1, trueSigma2),
    produceImage = "No"))


  # Incorrect ImageName format
  expect_error(mplnDataGenerator(nObservations = 50,
    dimensionality = 6,
    mixingProportions = c(0.79, 0.21),
    mu = rbind(trueMu1, trueMu2),
    sigma = rbind(trueSigma1, trueSigma2),
    produceImage = "Yes",
    ImageName = 1234))

})
# [END]