tests/testthat/test_normal_cases.R

In dr4pl: Dose Response Data Analysis using the 4 Parameter Logistic (4pl) Model

# -----------------------------------------------------------------------------
### Load libraries
#
library(drc)
library(testthat)

# -----------------------------------------------------------------------------
### User-defined functions
#
Fit4PLVariousOptions <- function(data.input,
                                 var.dose,
                                 var.response,
                                 var.ref = NULL) {

  types.method.init <- c("logistic", "Mead")
  types.method.optim <- c("Nelder-Mead", "BFGS", "CG", "SANN")

  data.input <- na.omit(data.input)  # Omit NA values from data

  if(length(var.ref) == 0) {

    data.whole <- subset(x = data.input, select = c(var.dose, var.response))
    colnames(data.whole) <- c("Dose", "Response")

    data.part <- data.whole

    for(method.init in types.method.init) {

      for(method.optim in types.method.optim) {

        expect_error(dr4pl(Response ~ Dose, data = !!data.part, method.init = !!method.init, method.optim = !!method.optim), NA)

      }
    }

  } else {

    data.whole <- subset(x = data.input, select = c(var.ref, var.dose, var.response))
    colnames(data.whole) <- c("Ref", "Dose", "Response")
    data.whole$Ref <- as.factor(data.whole$Ref)
    data.whole <- subset(x = data.whole, subset = Ref != "999")
    data.whole <- droplevels(data.whole)

    levels.ref <- levels(data.whole$Ref)

    for(i in 1:length(levels.ref)) {

      data.part <- subset(x = data.whole,
                          select = c(Dose, Response),
                          subset = Ref == levels.ref[i])

      for(method.init in types.method.init) {

        for(method.optim in types.method.optim) {

          expect_error(dr4pl(Response ~ Dose, data = !!data.part, method.init = !!method.init, method.optim = !!method.optim), NA)
        }
      }
    }
  }
}

TestNormalCases <- function() {

  Fit4PLVariousOptions(data.input = acidiq,
                       var.ref = "pct",
                       var.dose = "dose",
                       var.response = "rgr")

  Fit4PLVariousOptions(data.input = algae,
                       var.dose = "conc",
                       var.response = "vol")

  Fit4PLVariousOptions(data.input = etmotc,
                       var.ref = "pct1",
                       var.dose = "dose1",
                       var.response = "rgr1")

  Fit4PLVariousOptions(data.input = G.aparine,
                       var.dose = "dose",
                       var.response = "drymatter")

  Fit4PLVariousOptions(data.input = glymet,
                       var.ref = "pct",
                       var.dose = "dose",
                       var.response = "rgr")

  Fit4PLVariousOptions(data.input = heartrate,
                       var.dose = "pressure",
                       var.response = "rate")

  Fit4PLVariousOptions(data.input = leaflength,
                       var.dose = "Dose",
                       var.response = "DW")

  Fit4PLVariousOptions(data.input = lepidium,
                       var.dose = "conc",
                       var.response = "weight")

  Fit4PLVariousOptions(data.input = lettuce,
                       var.dose = "conc",
                       var.response = "weight")

  Fit4PLVariousOptions(data.input = mecter,
                       var.ref = "pct",
                       var.dose = "dose",
                       var.response = "rgr")

  Fit4PLVariousOptions(data.input = M.bahia,
                       var.dose = "conc",
                       var.response = "dryweight")

  Fit4PLVariousOptions(data.input = nasturtium,
                       var.dose = "conc",
                       var.response = "weight")

  Fit4PLVariousOptions(data.input = O.mykiss,
                       var.dose = "conc",
                       var.response = "weight")

  Fit4PLVariousOptions(data.input = P.promelas,
                       var.dose = "conc",
                       var.response = "dryweight")

  Fit4PLVariousOptions(data.input = ryegrass,
                       var.dose = "conc",
                       var.response = "rootl")

  Fit4PLVariousOptions(data.input = S.alba,
                       var.ref = "Herbicide",
                       var.dose = "Dose",
                       var.response = "DryMatter")

  Fit4PLVariousOptions(data.input = S.capricornutum,
                       var.dose = "conc",
                       var.response = "count")

  Fit4PLVariousOptions(data.input = secalonic,
                       var.dose = "dose",
                       var.response = "rootl")

  Fit4PLVariousOptions(data.input = spinach,
                       var.ref = "CURVE",
                       var.dose = "DOSE",
                       var.response = "SLOPE")

  Fit4PLVariousOptions(data.input = terbuthylazin,
                       var.dose = "dose",
                       var.response = "rgr")

  Fit4PLVariousOptions(data.input = vinclozolin,
                       var.ref = "exper",
                       var.dose = "conc",
                       var.response = "effect")
}

# --------------------------------------------------------------------------------
### Test
#
# context("Test whether running `dr4pl' on the data sets of `drc' using different
#         parameter settings does not draw any error")

test_that("Running `dr4pl' on the data sets of `drc' using different
           parameter settings should not draw any error.", {

  expect_error(TestNormalCases(), NA)
})