tests/testthat/test-full.R

In rIACI: Iberian Actuarial Climate Index Calculations

library(testthat)
library(rIACI)
library(reticulate)

# Python virtual environment
test_that("Set up Python virtual environment", {
  skip_on_cran()
  skip_if_not_installed("reticulate")

  if (!py_available(initialize = FALSE)) {
    skip("No Python available in this environment, skipping Python tests.")
  }

  env_dir <- file.path(tempdir(), "test_py_env")

  if (!dir.exists(env_dir)) {
    message("Creating Python virtual environment at: ", env_dir)
    virtualenv_create(env_dir)
    message("Installing Python packages...")
    py_install(
      packages = c("xarray", "pandas", "numpy", "netCDF4", "dask"),
      envname  = env_dir
    )
  }

  use_virtualenv(env_dir, required = TRUE)

  xarray <- import("xarray", convert = FALSE)
  expect_true(!is.null(xarray), "xarray should be available after installation.")
})

### 1: download_data
test_that("Integrated Test: download_data", {
  skip_on_cran()
  if (tolower(Sys.info()[["sysname"]]) == "linux" && Sys.getenv("GITHUB_ACTIONS") == "true") {
    skip("Skipping download test on Linux GitHub Actions due to keyring creation issues.")
  }

  # Define parameters for the download request
  start_year <- 2020
  end_year <- 2020
  start_month <- 1
  end_month <- 1
  variables <- c("2m_temperature", "total_precipitation")
  dataset <- "reanalysis-era5-land"
  area <- c(44, -10, 36, 4)

  # Create a temporary directory for downloaded data
  output_dir <- tempfile("cds_data_")
  dir.create(output_dir)

  # Retrieve ECMWF credentials from environment variables
  user_id <- Sys.getenv("ECMWF_USER_ID")
  user_key <- Sys.getenv("ECMWF_USER_KEY")

  # Skip the test if credentials are not set
  if (user_id == "" || user_key == "") {
    skip("ECMWF_USER_ID or ECMWF_USER_KEY not set, skipping download test")
  }

  # Call the download_data function to request data download
  download_data(
    start_year = start_year,
    end_year = end_year,
    start_month = start_month,
    end_month = end_month,
    variables = variables,
    dataset = dataset,
    area = area,
    output_dir = output_dir,
    user_id = user_id,
    user_key = user_key,
    max_retries = 1,
    retry_delay = 0,
    timeout = 7200
  )

  # Check that the downloaded file exists (either a .zip or .nc file)
  expected_nc  <- file.path(output_dir, "2020_01.nc")
  expect_true(file.exists(expected_nc))
})

### 2: process_data
test_that("Integrated Test: process_data", {
  skip_on_cran()
  skip_if_not_installed("reticulate")
  skip_if(!py_available(initialize = FALSE), "No Python available, skipping...")

  input_dir <- system.file("extdata", "testdata", package = "rIACI")
  expect_true(nchar(input_dir) > 0, info = "Cannot locate the 'testdata' directory in extdata")

  # Create a temporary directory for processed data output
  output_dir <- tempfile("processed_data_")
  dir.create(output_dir)

  # Call the process_data function to process the test data
  process_data(input_dir = input_dir, output_dir = output_dir, save_merged = FALSE)

  # Check that the processed data output directory contains files
  output_files <- list.files(output_dir)
  expect_gt(length(output_files), 0)
})

#3. export_data_to_csv
test_that("export_data_to_csv() with Python (skip on CRAN)", {
  skip_on_cran()
  skip_if_not_installed("reticulate")
  skip_if(!py_available(initialize = FALSE), "No Python available, skipping...")

  netcdf_file <- system.file("extdata", "testdata.nc", package = "rIACI")
  expect_true(file.exists(netcdf_file), info = "NetCDF file does not exist. Check path.")

  csv_output_dir <- tempfile("csv_output")
  dir.create(csv_output_dir)

    expect_error(
    export_data_to_csv(nc_file = netcdf_file, output_dir = csv_output_dir),
    NA
  )

  csv_files <- list.files(csv_output_dir, pattern = "\\.csv$", full.names = TRUE)
  expect_true(length(csv_files) > 0, info = "No CSV files were generated by export_data_to_csv.")

  required_cols <- c("TMAX", "TMIN", "PRCP", "WP", "time")
  for (file in csv_files) {
    df <- read.csv(file)
    missing_cols <- setdiff(required_cols, names(df))
    expect_true(length(missing_cols) == 0,
                info = paste("File", basename(file), "is missing columns:",
                             paste(missing_cols, collapse = ", ")))
  }
})

#4. Output all
test_that("Integrated test: output_all (monthly & seasonal)", {
  csv_dir <- system.file("extdata", "test_output", package = "rIACI")
  expect_true(dir.exists(csv_dir),
              info = "Directory 'test_output' not found in inst/extdata")

  original_csv_files <- list.files(csv_dir, pattern = "\\.csv$", full.names = TRUE)
  expect_true(length(original_csv_files) > 0,
              info = "No CSV files found in 'test_output' directory")

  csv_output_dir <- tempfile("csv_output")
  dir.create(csv_output_dir)
  file.copy(original_csv_files, csv_output_dir)

  copied_csv_files <- list.files(csv_output_dir, pattern = "\\.csv$", full.names = TRUE)
  expect_true(length(copied_csv_files) > 0,
              info = "No CSV files were copied to the temporary directory.")

  required_cols <- c("TMAX", "TMIN", "PRCP", "WP", "time")
  for (csvfile in copied_csv_files) {
    df <- read.csv(csvfile)
    missing_cols <- setdiff(required_cols, names(df))
    expect_true(length(missing_cols) == 0,
                info = paste("File", basename(csvfile),
                             "is missing columns:", paste(missing_cols, collapse = ", ")))
  }

  sea_dates <- seq(as.Date("1960-01-01"), as.Date("2023-12-01"), by = "month")
  sea_dates <- format(sea_dates, "%Y-%m")
  sea_values <- rep(NA, length(sea_dates))
  si <- sea_input(Date = sea_dates, Value = sea_values)

  # -----------------------
  #   a) Monthly
  # -----------------------
  monthly_output_dir_parent <- tempfile("iaci_monthly_results")
  expect_error(
    output_all(
      si          = si,
      input_dir   = csv_output_dir,
      output_dir  = monthly_output_dir_parent,
      freq        = "monthly",
      base.range  = c(1961, 1990),
      time.span   = c(1961, 2023)
    ),
    NA
  )
  monthly_output_dir <- file.path(monthly_output_dir_parent, "monthly")
  expect_true(dir.exists(monthly_output_dir),
              info = "output_all did not create a 'monthly' directory.")

  monthly_files <- list.files(monthly_output_dir, pattern = "\\.csv$")
  expect_true(length(monthly_files) > 0,
              info = "No CSV files generated in the monthly output directory.")

  sample_monthly <- read.csv(file.path(monthly_output_dir, monthly_files[1]))
  expect_true("IACI" %in% names(sample_monthly),
              info = "Monthly output CSV does not contain 'IACI' column.")

  # -----------------------
  #   b) Seasonal
  # -----------------------
  seasonal_output_dir_parent <- tempfile("iaci_seasonal_results")
  expect_error(
    output_all(
      si          = si,
      input_dir   = csv_output_dir,    # ????????? CSV ???
      output_dir  = seasonal_output_dir_parent,
      freq        = "seasonal",
      base.range  = c(1961, 1990),
      time.span   = c(1961, 2023)
    ),
    NA
  )
  seasonal_output_dir <- file.path(seasonal_output_dir_parent, "seasonal")
  expect_true(dir.exists(seasonal_output_dir),
              info = "output_all did not create a 'seasonal' directory.")

  seasonal_files <- list.files(seasonal_output_dir, pattern = "\\.csv$")
  expect_true(length(seasonal_files) > 0,
              info = "No CSV files generated in the seasonal output directory.")

  sample_seasonal <- read.csv(file.path(seasonal_output_dir, seasonal_files[1]))
  expect_true("IACI" %in% names(sample_seasonal),
              info = "Seasonal output CSV does not contain 'IACI' column.")
})


#5. csv_to_nc

test_that("csv_to_netcdf() [Python call]", {
  skip_on_cran()
  skip_if_not_installed("reticulate")
  skip_if(!py_available(initialize = FALSE), "No Python available, skipping...")

  csv_dir <- system.file("extdata", "testcsv", package = "rIACI")
  expect_true(dir.exists(csv_dir),
              info = "Directory inst/extdata/testcsv not found or not installed with the package.")

  merged_netcdf <- tempfile(fileext = ".nc")

  expect_error(csv_to_netcdf(csv_dir = csv_dir,  output_file = merged_netcdf,
                             freq = "monthly"),NA)


  expect_true(file.exists(merged_netcdf),
              info = "csv_to_netcdf did not generate the merged NetCDF file.")
})