tests/testthat/test-MAP.R

In MAP: Multimodal Automated Phenotyping

# devtools::load_all()
# library(MAP)

context('MAP')

set.seed(123)
n = 400
ICD = c(rpois(n / 4, 10), rpois(n / 4, 1), rep(0, n / 2))
NLP = c(rpois(n / 4, 10), rpois(n / 4, 1), rep(0, n / 2))
mat = Matrix(data = cbind(ICD, NLP), sparse = TRUE)
note = Matrix(rpois(n, 10) + 5, ncol = 1, sparse = TRUE)

res_full = MAP(mat = mat, note = note, verbose = FALSE)
clusters = res_full$scores[, 1] > res_full$cut.MAP

test_MAP <- function() {

  expect_equal(dim(head(res_full$scores)), c(6, 1))

  expect_equal(round(res_full$cut.MAP, 3), 0.125)

  expect_equal(sum(clusters), 102)

}
test_that('MAP', test_MAP())

test_fitproj <- function() {

  # take 50%
  set.seed(1)
  res_sample = MAP(mat = mat, note = note, subset_sample = TRUE,
                   subset_sample_size = 82, verbose = FALSE)

  expect_equal(round(res_sample$cut.MAP, 3), 0.292)

  clusters_sample = res_sample$scores[, 1] > res_sample$cut.MAP
  expect_equal(sum(clusters_sample), 101)

  roc_obj = pROC::roc(clusters, as.numeric(clusters_sample))
  expect_equal(round(roc_obj$auc, 3), 0.995)

  # take 10%
  set.seed(1)

  res_sample = tryCatch(expr = MAP(mat = mat, note = note, subset_sample = TRUE,
                                   subset_sample_size = 16, verbose = FALSE),
                            warning = function(w_msg) {
                              if (length(as.character(w_msg)) > 1 ||
                                  !grepl('The clustering step does not converge for variable ICD_log', w_msg)) {
                                print(w_msg)
                              }
                              suppressWarnings({
                                  MAP(mat = mat, note = note, subset_sample = TRUE,
                                                   subset_sample_size = 16, verbose = FALSE)
                                })
                            })

  expect_equal(round(res_sample$cut.MAP, 3), 0.148)

  clusters_sample = res_sample$scores[, 1] > res_sample$cut.MAP
  expect_equal(sum(clusters_sample), 101)

  roc_obj = pROC::roc(clusters, as.numeric(clusters_sample))
  expect_equal(round(roc_obj$auc, 3), 0.995)
}
test_that('fitproj', test_fitproj())