tests/testthat/testGE.R

In baycn: Bayesian Inference for Causal Networks

context('data: GE')

test_that('mhEdge infers the correct graph',{

  # Adjacency matrix for M1 and M2 ---------------------------------------------

  # Adjacency matrix for M1
  am_m1 <- matrix(c(0, 1, 0,
                    0, 0, 1,
                    0, 0, 0),
                  byrow = TRUE,
                  nrow = 3)

  # Adjacency matrix for M2
  am_m2 <- matrix(c(0, 1, 0,
                    0, 0, 0,
                    0, 1, 0),
                  byrow = TRUE,
                  nrow = 3)

  # Simulate data for M1 and M2 ------------------------------------------------

  set.seed(5)

  data_m1 <- simdata(b0 = 0,
                     N = 500,
                     s = 1,
                     ss = 1,
                     graph = 'm1_ge')

  data_m2 <- simdata(b0 = 0,
                     N = 500,
                     s = 1,
                     ss = 1,
                     graph = 'm2_ge')

  # Run baycn on M1 and M2 -----------------------------------------------------

  baycn_m1 <- mhEdge(adjMatrix = am_m1,
                     burnIn = 0.2,
                     data = data_m1,
                     iterations = 1000,
                     nCPh = 0,
                     nGV = 0,
                     pmr = FALSE,
                     prior = c(0.05,
                               0.05,
                               0.9),
                     progress = TRUE,
                     thinTo = 200)

  baycn_m2 <- mhEdge(adjMatrix = am_m2,
                     burnIn = 0.2,
                     data = data_m2,
                     iterations = 1000,
                     nCPh = 0,
                     nGV = 0,
                     pmr = FALSE,
                     prior = c(0.05,
                               0.05,
                               0.9),
                     progress = FALSE,
                     thinTo = 200)

  # Calculate the MSE for M1 and M2 --------------------------------------------

  # Expected probabilities for M1
  ep_m1 <- matrix(c(1/3, 2/3, 0,
                    2/3, 1/3, 0),
                  byrow = TRUE,
                  nrow = 2)

  mse_m1 <- sum((baycn_m1@posteriorES[, 2:4] - ep_m1)^2)

  # Expected probabilities for M2
  ep_m2 <- matrix(c(1, 0, 0,
                    0, 1, 0),
                  byrow = TRUE,
                  nrow = 2)

  mse_m2 <- sum((baycn_m2@posteriorES[, 2:4] - ep_m2)^2)

  expect_true(mse_m1 < 0.1)
  expect_true(mse_m2 < 0.1)

})