tests/testthat/test_time_series.R

In Techtonique/crossval: Generic cross-validation functions

context("Test on time series models")


library(testthat)
library(crossval)
library(forecast)


# Example 1 -----

res1 <- crossval_ts(y=AirPassengers, initial_window = 10, fcast_func = thetaf)


# Example 2 -----

fcast_func <- function (y, h, ...)
{
      forecast::forecast(forecast::auto.arima(y, ...),
      h=h, ...)
}

res2 <- crossval_ts(y=AirPassengers, initial_window = 10, fcast_func = fcast_func)



# Example 3 -----

fcast_func <- function (y, h, ...)
{
      forecast::forecast(forecast::ets(y, ...),
      h=h, ...)
}

res3 <- crossval_ts(y=AirPassengers, initial_window = 10, fcast_func = fcast_func)


# Example 4 -----

 set.seed(123)
 x <- ts(matrix(rnorm(50), nrow = 25))

 fcast_func <- function(y, h = 5, type_forecast=c("mean", "median"))
 {
  type_forecast <- match.arg(type_forecast)

  if (type_forecast == "mean")
  {
   means <- colMeans(y)
   return(list(mean = t(replicate(n = h, expr = means))))
  } else {
   medians <- apply(y, 2, median)
   return(list(mean = t(replicate(n = h, expr = medians))))
  }

 }

res4 <- crossval::crossval_ts(y = x, fcast_func = fcast_func, fit_params = list(type_forecast = "median"))

res5 <- crossval::crossval_ts(y = x, fcast_func = fcast_func, fit_params = list(type_forecast = "mean"))



# tests -----

mean1 <- colMeans(res1)
mean2 <- colMeans(res2)
mean3 <- colMeans(res3)
mean4 <- colMeans(res4)
mean5 <- colMeans(res5)

test_that("tests on mean RMSE and MAPE", {
  expect_equal(as.numeric(round(mean1["RMSE"], 2)), 51.43 )
  expect_equal(as.numeric(round(mean1["MAPE"], 2)), 0.16 )
  expect_equal(as.numeric(round(mean2["RMSE"], 2)), 51.91)
  expect_equal(as.numeric(round(mean2["MAPE"], 2)), 0.15)
  expect_equal(as.numeric(round(mean3["RMSE"], 2)), 45.34)
  expect_equal(as.numeric(round(mean3["MAPE"], 2)), 0.14)
  expect_equal(as.numeric(round(mean4["RMSE"], 2)), 0.97)
  expect_equal(as.numeric(round(mean5["MAPE"], 2)), 1.43)
})