tests/testthat/test_load_data.R

In gpw13/billionaiRe: Calculate the WHO Triple Billions

# WHDH --------------------------------------------------------------------

# TODO: Get the test to pass.
testthat::test_that("WHDH: a single indicator is correctly downloaded", {
  # wrangled_ind <- load_billion_data("wrangled_data", "hpop", "alcohol") %>%
  #   dplyr::pull(ind) %>%
  #   unique()
  #
  # projected_ind <- load_billion_data("projected_data", "hpop", "alcohol") %>%
  #   dplyr::pull(ind) %>%
  #   unique()
  #
  # testthat::expect_equal(wrangled_ind, "alcohol")
  # testthat::expect_equal(projected_ind, "alcohol")
})

# TODO: currently failing because of a bug in the data_pipeline
testthat::test_that("WHDH: load_billion_data downloads the right data_type", {
  # df_wrangled <- load_billion_data("wrangled_data", "hpop", "alcohol")
  # df_projected <- load_billion_data("projected_data", "hpop", "alcohol")

  # testthat::expect_equal(unique(df_wrangled$type), "estimated")
  # testthat::expect_true("projected" %in% unique(df_projected$type))
  # testthat::expect_equal(names(df_wrangled), xmart_cols("wrangled_data"))
  # testthat::expect_equal(names(df_projected), xmart_cols("projected_data"))
})

# TODO: Get the test to pass.
testthat::test_that("WHDH: uhc_espar is correctly downloaded (from the hep_espar folder)", {
  # c("wrangled_data", "projected_data") %>%
  #   purrr::walk(~ {
  #     df_hep <- load_billion_data(.x, "hep", "espar")
  #     df_uhc <- load_billion_data(.x, "uhc", "espar")
  #     testthat::expect_equal(df_hep, df_uhc)
  #   })
})

# TODO: Get the test to pass.
testthat::test_that("WHDH: load_billion_data correctly handles ind_codes = \"all\"", {
  # c("hep", "hpop", "uhc") %>%
  #   purrr::map(~ {
  #     df <- suppressMessages(load_billion_data("wrangled_data", .x, "all"))
  #     inds <- unique(df$ind) %>%
  #       # Remove sub-indicators like water_rural, water_urban, etc.
  #       stringr::str_replace("_rural|_urban|_num|_denom", "") %>%
  #       # Remove espar categories and sub-categories
  #       stringr::str_replace("espar.*", "espar") %>%
  #       unique() %>%
  #       sort()
  #     testthat::expect_equal(inds, unname(sort(get_valid_inds("wrangled_data", .x))))
  #   })
})

# TODO: Get the test to pass.
testthat::test_that("WHDH: load_billion_data correctly handles billion = \"all\"", {
  # df <- suppressMessages(load_billion_data("wrangled_data", "all"))
  #
  # inds <- unique(df$ind) %>%
  #   # Remove sub-indicators like water_rural, water_urban, etc.
  #   stringr::str_replace("_rural|_urban|_num|_denom", "") %>%
  #   # Remove espar categories and sub-categories
  #   stringr::str_replace("espar.*", "espar") %>%
  #   unique() %>%
  #   sort()
  #
  # valid_inds <- c("hep", "hpop", "uhc") %>%
  #   purrr::map(~ get_valid_inds("wrangled_data", .x)) %>%
  #   unlist() %>%
  #   unique() %>%
  #   unname() %>%
  #   sort()
  #
  # testthat::expect_equal(inds, valid_inds)
})

# TODO: Get the test to pass.
testthat::test_that("WHDH: load_billion_data correctly handles final_data", {
  # df <- suppressMessages(load_billion_data("final_data"))
  # df2 <- suppressMessages(load_billion_data("final_data", "all"))
  # df3 <- suppressMessages(load_billion_data("final_data", "all", "all"))
  #
  # inds <- sort(unique(df$ind))
  # expected_inds <- c("hep", "hpop", "uhc") %>%
  #   purrr::map(
  #     ~ billion_ind_codes(.x, include_covariates = FALSE, include_calculated = TRUE, include_subindicators = TRUE)
  #   ) %>%
  #   unlist() %>%
  #   unname() %>%
  #   # Removing espar subindicators because those are not in final_data
  #   stringr::str_replace("espar.*", "espar") %>%
  #   unique() %>%
  #   sort()
  #
  # testthat::expect_equal(df, df2)
  # testthat::expect_equal(df, df3)
  # # TODO: currently failing due to issues with data_pipeline
  # # testthat::expect_equal(inds, expected_inds)
})

# TODO: Get the test to pass.
testthat::test_that("WHDH: load_billion_data correctly handles na_rm", {
  # # Downloading all billions because some indicators might have a full timeseries
  # # Using all indicators ensures there's at least one rows with an NA value
  # df_na <- load_billion_data("wrangled_data", "all", na_rm = FALSE) %>%
  #   dplyr::filter(is.na(value)) %>%
  #   suppressMessages()
  # df_no_na <- load_billion_data("wrangled_data", "all", na_rm = TRUE) %>%
  #   dplyr::filter(is.na(value)) %>%
  #   suppressMessages()
  #
  # testthat::expect_true(nrow(df_na) > 0)
  # testthat::expect_true(nrow(df_no_na) == 0)
})

# load_misc_data ---------------------------------

testthat::test_that("load_misc_data works as expected", {
  testthat::expect_equal(105, nrow(load_misc_data("pulse/pulse.csv")))
  testthat::expect_equal(106, nrow(load_misc_data("pulse/pulse.csv", col_names = FALSE)))
})