tests/testthat/test-prep_data_utils.R

In healthcareai: Tools for Healthcare Machine Learning

context("Testing data_prep utilities")

# Setup ------------------------------------------------------------------------
# set seed for reproducibility
set.seed(7)
# build data set to predict whether or not animal_id is a is_ween
n <- 100
d <- data.frame(song_id = 1:n,
       length = rnorm(n, mean = 4, sd = 1),
       tuba_flag = sample(c(0, 1), size = n, replace = T),
       drum_flag = sample(c(0, 1, NA), size = n, replace = T),
       reaction = sample(c("Love", "Huh", "Dislike", "Mixed"),
                         size = n, replace = T),
       date_col = lubridate::ymd("2002-03-04") + lubridate::days(1:10),
       posixct_col = as.POSIXct(
         lubridate::ymd("2004-03-04") + lubridate::days(1:10)),
       col_DTS = as.character(
         lubridate::ymd("2006-03-04") + lubridate::days(1:10)),
       missing82 = sample(1:10, n, replace = TRUE),
       missing64 = sample(100:300, n, replace = TRUE),
       stringsAsFactors = FALSE)

d$missing82[sample(1:n, 82, replace = FALSE)] <- NA
d$missing64[sample(1:n, 64, replace = FALSE)] <- NA

pd <- prep_data(d, song_id, col_DTS, outcome = length)

# Tests ------------------------------------------------------------------------
test_that("find_0_1_cols returns numeric flag columns", {
  exp <- c("tuba_flag", "drum_flag")
  expect_equal(find_0_1_cols(d), exp)
})

test_that("find_mostly_missing_cols returns numeric flag columns", {
  expect_equal(find_mostly_missing_cols(d, 80), "missing82")

  exp <- c("missing64", "missing82")
  expect_equal(find_mostly_missing_cols(d, 60), exp)
})

test_that("find_date_cols returns date columns", {
  exp <- c("date_col", "posixct_col", "col_DTS")
  expect_equal(find_date_cols(d), exp)
})

test_that("check_rec_obj", {
  expect_s3_class(check_rec_obj(recipes::recipe(d)), "recipe")
  expect_s3_class(
    check_rec_obj(structure(d, "recipe" = recipes::recipe(d, drum_flag ~ .))),
    "recipe")
  expect_error(check_rec_obj(d))
  expect_error(check_rec_obj("recipe"))
})

test_that("find_new_missingness returns length-0 char if no new missingness", {
  expect_equal(0, length(find_new_missingness(d, attr(pd, "recipe"))))
})

test_that("find_new_missingness returns predictors with new missingness", {
  d$reaction[1] <- NA
  d$date_col[2:4] <- NA
  expect_true(all(c("reaction", "date_col") %in% find_new_missingness(d, attr(pd, "recipe"))))
})

test_that("find_new_missingness doesn't return ignore or outcome columns", {
  d$length[1:5] <- NA
  d$song_id <- NULL
  expect_equal(0, length(find_new_missingness(d, attr(pd, "recipe"))))
})

test_that("get_factor_levels works when just one factor", {
  fl <- get_factor_levels(d)
  expect_equal("list", class(fl))
  expect_equal("reaction", names(fl))
  expect_setequal(as.character(d$reaction), names(fl$reaction))
})

test_that("get_factor_levels works on character and factors", {
  pfl <- get_factor_levels(pima_diabetes)
  expect_equal("list", class(pfl))
  expect_equal(c("weight_class", "diabetes"), names(pfl))
  expect_setequal(as.character(pima_diabetes$weight_class), names(pfl$weight_class))
  expect_setequal(as.character(pima_diabetes$diabetes), names(pfl$diabetes))
})