tests/testthat/test-view_on_joint_distribution.R

In ffp: Fully Flexible Probabilities for Stress Testing and Portfolio Construction

ret_ts  <- diff(log(EuStockMarkets))
ret_mtx <- matrix(ret_ts, ncol = 4)
ret_xts <- xts::xts(ret_mtx, order.by = Sys.Date() + 1:nrow(ret_mtx))
ret_tbl <- tibble::as_tibble(ret_ts)

p <- rep(1 / nrow(ret_ts), nrow(ret_ts))
simul_marg <- bootstrap_scenarios(ret_mtx, as_ffp(p), as.double(NROW(p)))
simul_cop <- apply(simul_marg, 2, stats::pnorm) # assuming normal copula

vmean_ts <- view_on_joint_distribution(x = ret_ts, simul = simul_cop, p = p)
test_that("`view_on_joint_distribution` works for ts", {
  expect_type(vmean_ts, "list")
  expect_s3_class(vmean_ts, "ffp_views")
  expect_length(vmean_ts, 2L)
  expect_named(vmean_ts, c("Aeq", "beq"))
})

vmean_mtx <- view_on_joint_distribution(x = ret_mtx, simul = simul_cop, p = p)
test_that("`view_on_joint_distribution` works for matrix", {
  expect_type(vmean_mtx, "list")
  expect_s3_class(vmean_mtx, "ffp_views")
  expect_length(vmean_mtx, 2L)
  expect_named(vmean_mtx, c("Aeq", "beq"))
})

vmean_xts <- view_on_joint_distribution(x = ret_xts, simul = simul_cop, p = p)
test_that("`view_on_joint_distribution` works for xts", {
  expect_type(vmean_xts, "list")
  expect_s3_class(vmean_xts, "ffp_views")
  expect_length(vmean_xts, 2L)
  expect_named(vmean_xts, c("Aeq", "beq"))
})

vmean_tbl <- view_on_joint_distribution(x = ret_tbl, simul = simul_cop, p = p)
test_that("`view_on_joint_distribution` works for tbl_df", {
  expect_type(vmean_tbl, "list")
  expect_s3_class(vmean_tbl, "ffp_views")
  expect_length(vmean_tbl, 2L)
  expect_named(vmean_tbl, c("Aeq", "beq"))
})