tests/testthat/test-checkValidConfig.R

context("checkValidConfig testing")

test_that("proper warning is returned for (and only for) < 5 numeric target vals", {
  the.data4 <- data.frame(y = rep(1:4, 100))
  the.data5 <- data.frame(y = rep(1:5, 100))
  names <- list(y = "y")
  config <- list(cp = .01, model.algorithm = "rpart")
  expect_identical(
    checkValidConfig(config, the.data4, names)[1],
    TRUE
  )
  # Expect no errors, warnings, or messages
  expect_equal(
    sum(checkValidConfig(config, the.data5, names)),
    0
  )
})

test_that("proper error throwing is done for invalid control parameter", {
  the.data <- data.frame(y = 1:100)
  names <- list(y = "y")
  config0 <- list(cp = -0.1, model.algorithm = "rpart")
  config1 <- list(cp = 1.1, model.algorithm = "rpart")
  expect_identical(
    checkValidConfig(config0, the.data, names)[2],
    TRUE
  )
  expect_identical(
    checkValidConfig(config1, the.data, names)[2],
    TRUE
  )
})
alteryx/AlteryxPredictive documentation built on May 12, 2019, 1:37 a.m.