test_that("S2MVarSelection works", {
  # Data Generation

  n <- 100
  p <- 20
  X <- matrix(rnorm(n * p), n, p)
  beta <- exp(rnorm(p))
  Y <- X %*% beta + rnorm(n, 0, 1)
  df <- data.frame(X, Y)
  rv.hs <- bayesreg::bayesreg(Y ~ ., df, "gaussian", "horseshoe+", 150, 100)
  Beta <- t(rv.hs$beta)
  H <- 12

  #--------------------- S2MVarSelection -----------------------#

  # Check for NULL or NaN values of Beta
  testthat::expect_error(S2MVarSelection(NULL, H))

  # Check for matrix data type of Beta
  testthat::expect_error(S2MVarSelection(Beta[, 1], H))

  # Check for NULL or NaN values for H
  testthat::expect_error(S2MVarSelection(Beta, NULL))

  # Check for numeric data type of H
  testthat::expect_error(S2MVarSelection(Beta, "H"))

  # Expecting output to be of predefined length for impVariablesGLM
  testthat::expect_length((S2MVarSelection(Beta, H)), H)

  # Check for H must be less than total covariates.
  testthat::expect_error(S2MVarSelection(Beta, p + 50))

  #--------------------- S2MVarSelectionV1 ---------------------#

  # the medians of the absolute values of the posterior samples of each Beta vector
  abs.post.median <- seq(0, length = p)
  for (i in 1:p) {
    abs.post.median[i] <- stats::median(abs(Beta[, i]))
  }

  S2M <- list(p = p, abs.post.median = abs.post.median)

  # Check for NULL or NaN values of abs.post.median
  testthat::expect_error(S2MVarSelectionV1(NULL, H))

  # Check for vector data type of abs.post.median
  testthat::expect_error(S2MVarSelectionV1(list(abs.post.median = "ABC"), H))

  # Check for NULL or NaN values for H
  testthat::expect_error(S2MVarSelectionV1(S2M, NULL))

  # Check for numeric data type of H
  testthat::expect_error(S2MVarSelectionV1(S2M, "H"))

  # Check for non negative values in abs.post.median
  testthat::expect_error(S2MVarSelectionV1(S2M = list(abs.post.median = c(-1, -2, 3)), "H"))

  # Check for H must be less than total covariates.
  testthat::expect_error(S2MVarSelection(Beta, p + 50))

  # Check for vector data type of abs.post.median
  testthat::expect_length(S2MVarSelectionV1(S2M, H), H)

  #--------------------- OptimalHbi ---------------------#

  # Check for NULL or NaN values in b.i
  testthat::expect_error(OptimalHbi(NULL, Hbi = c(17, 12, 12, 12, 12, 6, 6, 6, 4, 5)))

  # Check for vector data type of b.i
  testthat::expect_error(OptimalHbi(matrix(rnorm(4), 2, 2), Hbi = c(17, 12, 12, 12, 12, 6, 6, 6, 4, 5)))

  # Check for the length of b.i and H.b.i
  testthat::expect_error(OptimalHbi(c(1, 2, 4), Hbi = c(17, 12, 12, 12, 12, 6, 6, 6, 4, 5)))
})