tests/testthat/test_thin_data.R

In pratikunterwegs/atlastools: Tools for Pre-Processing High-Throughput Animal Tracking Data

context("thin cleaned data\n")
testthat::test_that("aggregating cleaned data", {
  interval <- 60
  # make test_data
  test_data <- data.table::data.table(
    `location-lat` = as.double(seq(1000)), # mimic movebank naming
    y = as.double(seq(1000)),
    time = as.numeric(seq(1000)),
    ts = as.POSIXct(seq(1000),
      origin = "1970-01-01"
    ),
    id = as.factor("abc"),
    VARX = runif(1000) + 300,
    VARY = runif(1000) + 300,
    COVXY = runif(1000) + 200
  )

  # make SD
  test_data[, SD := sqrt(VARX + VARY + 2 * COVXY)]

  # run function --- expect warning
  # expect a warning for the missing 'x' column name
  testthat::expect_warning(
    test_output <- atl_thin_data(test_data,
      id_columns = "id",
      interval = 60,
      time = "time",
      method = "aggregate"
    )
  )

  # do tests
  # test that the vector class is data.table and data.frame
  testthat::expect_s3_class(
    object = test_output,
    class = c("data.table", "data.frame")
  )

  # check that some rows are removed or that none are added
  testthat::expect_gte(nrow(test_data), nrow(test_output))

  # check that the count is made
  atlastools:::atl_check_data(test_output,
    names_expected = c("count", "SD", "VARX")
  )

  # check there is no SD column
  testthat::expect_error(
    atlastools:::atl_check_data(test_output, names_expected = c("COVXY"))
  )
})

# test data without COVX and COVY
testthat::test_that("aggregating cleaned data", {
  interval <- 60
  # make test_data
  test_data <- data.table::data.table(
    x = as.double(1:1e3),
    y = as.double(1:1e3),
    time = as.numeric(1:1e3),
    ts = as.POSIXct(1:1e3,
      origin = "1970-01-01"
    ),
    id = as.factor("abc")
  )

  # run function
  test_output <- atl_thin_data(test_data,
    id_columns = "id",
    interval = 60,
    method = "aggregate"
  )

  # do tests
  # test that the vector class is data.table and data.frame
  testthat::expect_s3_class(
    object = test_output,
    class = c("data.table", "data.frame")
  )

  # check that some rows are removed or that none are added
  testthat::expect_gte(nrow(test_data), nrow(test_output))

  # check that the count is made
  atlastools:::atl_check_data(test_output, names_expected = "count")

  # check there is no SD column
  testthat::expect_error(
    atlastools:::atl_check_data(test_output, names_expected = c("VARX", "SD"))
  )
})

# test for resampling
testthat::test_that("subsampling cleaned data", {
  interval <- 60
  # make test_data
  test_data <- data.table::data.table(
    x = as.double(1:1e3),
    y = as.double(1:1e3),
    time = as.numeric(1:1e3),
    ts = as.POSIXct(1:1e3,
      origin = "1970-01-01"
    ),
    id = as.factor("abc"),
    VARX = runif(1000) + 300,
    VARY = runif(1000) + 300,
    COVXY = runif(1000) + 200
  )

  # make SD
  test_data[, SD := sqrt(VARX + VARY + 2 * COVXY)]

  # run function
  test_output <- atlastools::atl_thin_data(test_data,
    interval = 60,
    id_columns = "id",
    method = "subsample"
  )

  # do tests
  # test that the vector class is data.table and data.frame
  testthat::expect_s3_class(
    object = test_output,
    class = c("data.table", "data.frame")
  )

  # check that some rows are removed or that none are added
  testthat::expect_gte(nrow(test_data), nrow(test_output))

  # check that the count is made
  atlastools:::atl_check_data(test_output, names_expected = "count")

  # check there is no SD column
  testthat::expect_silent(
    atlastools:::atl_check_data(test_output, names_expected = c("COVXY", "SD"))
  )
})

# test for other option
testthat::test_that("other options fail", {
  interval <- 60
  # make test_data
  test_data <- data.table::data.table(
    x = as.double(1:1e3),
    y = as.double(1:1e3),
    time = as.numeric(1:1e3),
    ts = as.POSIXct(1:1e3,
      origin = "1970-01-01"
    ),
    id = as.factor("abc"),
    VARX = runif(1000) + 300,
    VARY = runif(1000) + 300,
    COVXY = runif(1000) + 200
  )
  # run function
  testthat::expect_error(atlastools::atl_thin_data(test_data,
    interval = 60,
    id_columns = "id",
    method = "another_option"
  ))
})