if (identical(Sys.getenv("NOT_CRAN", unset = "true"), "true")) {
  test_that("Test normalization", {
    gimap_dataset <- get_example_data("gimap") %>%
      gimap_filter() %>%
      gimap_annotate(cell_line = "HELA") %>%
      gimap_normalize(
        timepoints = "day"
      )

    # make sure the important columns are there
    testthat::expect_true(
      all(c("target_type", "lfc", "rep", "crispr_score", "unexpressed_ctrl_flag")
      %in% colnames(gimap_dataset$normalized_log_fc))
    )

    neg_controls <- gimap_dataset$normalized_log_fc %>%
      dplyr::filter(norm_ctrl_flag == "negative_control") %>%
      dplyr::group_by(rep) %>%
      dplyr::summarize(neg_ctrl_med = median(crispr_score)) %>%
      dplyr::pull(neg_ctrl_med)

    # We expect negative controls to be now equal to 0
    testthat::expect_equal(neg_controls[2:4], c(0, 0, 0))

    pos_controls <- gimap_dataset$normalized_log_fc %>%
      dplyr::filter(norm_ctrl_flag == "positive_control") %>%
      dplyr::group_by(rep) %>%
      dplyr::summarize(pos_ctrl_med = median(crispr_score)) %>%
      dplyr::pull(pos_ctrl_med)

    # We expect positive controls to be now equal to -1
    testthat::expect_equal(
      round(pos_controls[-1]),
      round(c(-1, -1, -1))
    )
  })
}