tests/testthat/test.analysis.R

In llama: Leveraging Learning to Automatically Manage Algorithms

test_that("contributions returns contributions", {
    d = list(data=data.frame(a=c(1,2,3), b=c(4,2,1)), performance=c("a", "b"), minimize=TRUE)
    class(d) = "llama.data"
    cs = contributions(d)
    expect_equal(cs[[1]], -1.5)
    expect_equal(cs[[2]], -2.5)
})

test_that("contributions respects minimize", {
    d = list(data=data.frame(a=c(1,2,3), b=c(4,2,1)), performance=c("a", "b"), minimize=FALSE)
    class(d) = "llama.data"
    cs = contributions(d)
    expect_equal(cs[[1]], 4)
    expect_equal(cs[[2]], 5)
})

test_that("contributions raises error if no data given", {
    expect_error(contributions(), "Need data to determine contributions!")
})