context("Testing midas_r methods")
set.seed(1001)
n <- 250
trend <- c(1:n)
x <- rnorm(4 * n)
z <- rnorm(12 * n)
fn_x <- nealmon(p = c(1, -0.5), d = 8)
fn_z <- nealmon(p = c(2, 0.5, -0.1), d = 17)
y <- 2 + 0.1 * trend + mls(x, 0:7, 4) %*% fn_x + mls(z, 0:16, 12) %*% fn_z + rnorm(n)


spd <- split_data(list(y = y, x = x, trend = trend, z = z), 1:200, 201:250)

accuracy <- sqrt(.Machine$double.eps)

## Add test for preserving zoo and other attributes.
## Add test for printing out the estimation sample start and end.


test_that("midas_r preserves ts attribute", {
  spd$indata$y <- ts(spd$indata$y, start = 1, frequency = 1)
  a <- midas_r(y ~ trend + mls(x, 0:7, 4, nealmon) + mls(z, 0:16, 12, nealmon), start = list(x = c(1, -0.5), z = c(2, 0.5, -0.1)), data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  expect_true(inherits(fa$mean, "ts"))
  expect_true(identical(time(fa$mean), time(ts(201:250, start = 201))))

  spd$indata$y <- ts(spd$indata$y, start = 1950, frequency = 4)
  a <- midas_r(y ~ trend + mls(x, 0:7, 4, nealmon) + mls(z, 0:16, 12, nealmon), start = list(x = c(1, -0.5), z = c(2, 0.5, -0.1)), data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  expect_true(inherits(fa$mean, "ts"))
  expect_true(identical(time(fa$mean), time(ts(201:250, start = c(2000, 1), frequency = 4))))

  spd$indata$y <- ts(spd$indata$y, start = 1990, frequency = 12)
  a <- midas_r(y ~ trend + mls(x, 0:7, 4, nealmon) + mls(z, 0:16, 12, nealmon), start = list(x = c(1, -0.5), z = c(2, 0.5, -0.1)), data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  expect_true(inherits(fa$mean, "ts"))
  expect_lt(sum(abs(time(fa$mean) - time(ts(201:250, start = c(2006, 9), frequency = 12)))), accuracy)
})

test_that("midas_u preserves ts attribute", {
  spd$indata$y <- ts(spd$indata$y, start = 1, frequency = 1)
  a <- midas_u(y ~ trend + mls(x, 0:7, 4) + mls(z, 0:16, 12), data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  expect_true(inherits(fa$mean, "ts"))
  expect_true(identical(time(fa$mean), time(ts(201:250, start = 201))))

  spd$indata$y <- ts(spd$indata$y, start = 1950, frequency = 4)
  a <- midas_r(y ~ trend + mls(x, 0:7, 4, nealmon) + mls(z, 0:16, 12, nealmon), start = list(x = c(1, -0.5), z = c(2, 0.5, -0.1)), data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  expect_true(inherits(fa$mean, "ts"))
  expect_true(identical(time(fa$mean), time(ts(201:250, start = c(2000, 1), frequency = 4))))

  spd$indata$y <- ts(spd$indata$y, start = 1990, frequency = 12)
  a <- midas_r(y ~ trend + mls(x, 0:7, 4, nealmon) + mls(z, 0:16, 12, nealmon), start = list(x = c(1, -0.5), z = c(2, 0.5, -0.1)), data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  expect_true(inherits(fa$mean, "ts"))
  expect_lt(sum(abs(time(fa$mean) - time(ts(201:250, start = c(2006, 9), frequency = 12)))), accuracy)
})


test_that("forecast for midas_r and midas_u is the same", {
  a <- midas_r(y ~ trend + mls(x, 0:7, 4) + mls(z, 0:16, 12), start = NULL, data = spd$indata)
  b <- midas_u(y ~ trend + mls(x, 0:7, 4) + mls(z, 0:16, 12), start = NULL, data = spd$indata)
  fa <- forecast(a, newdata = spd$outdata)
  fb <- forecast(b, newdata = spd$outdata)
  expect_lt(sum(abs(fa$mean - fb$mean)), accuracy)
})