tests/testthat/test_handle_NA.r

In Tenenhaus/RGCCA: Regularized and Sparse Generalized Canonical Correlation Analysis for Multiblock Data

# '# Test handle_NA
#
# '''
# Load data
data(Russett)
blocks <- list(
  agriculture = Russett[, seq(3)],
  industry = Russett[, 4:5],
  politic = Russett[, 6:11],
  target = Russett[, 11]
)

# Add missing values
blocks[[1]][c(2, 4, 8), ] <- NA
blocks[[2]][c(12, 23), 1] <- NA
blocks[[2]][17, 2] <- NA
blocks[[3]][c(30, 32), 3] <- NA
blocks[[3]][40, ] <- NA
blocks[[4]][42] <- NA
ind_NA <- c(2, 4, 8, 12, 17, 23, 30, 32, 40, 42)

test_that("handle_NA selects the common rows without missing values when
  NA_method is \"na.omit\"", {
  tmp <- handle_NA(blocks, NA_method = "na.omit")
  for (j in seq_along(blocks)) {
    expect_equal(tmp$blocks[[j]], subset_rows(blocks[[j]], -ind_NA))
    expect_false(tmp$na.rm)
  }
})
test_that("handle_NA raises an error if there is less than 3 subjects
          left after removing missing values", {
  bad_blocks <- blocks
  bad_blocks[[2]][-c(1, 3), 1] <- NA
  expect_error(handle_NA(bad_blocks, NA_method = "na.omit"),
    paste0(
      "Less than 3 subjects have no missing values, ",
      "choose another missing value handling method ",
      "or work on your dataset."
    ),
    fixed = TRUE
  )
})
test_that("handle_NA leaves the blocks untouched when NA_method is
          \"na.ignore\"", {
  tmp <- handle_NA(blocks, NA_method = "na.ignore")
  expect_equal(tmp$blocks, blocks)
  expect_true(tmp$na.rm)
})