Added test file for func_MCMC_graph_cpp() (#11)

tests/testthat/test-func_MCMC_graph_cpp.R

@@ -0,0 +1,38 @@
+# Load the example dataset
+dataset <- list(
+  "X" = simData[[1]]$X,
+  "t" = simData[[1]]$time,
+  "di" = simData[[1]]$status
+)
+
+# Run a Bayesian Cox model
+
+## Initial value: null model without covariates
+initial <- list("gamma.ini" = rep(0, ncol(dataset$X)))
+
+# Prior parameters
+hyperparPooled <- list(
+  "c0"     = 2, # prior of baseline hazard
+  "tau"    = 0.0375, # sd (spike) for coefficient prior
+  "cb"     = 20, # sd (slab) for coefficient prior
+  "pi.ga"  = 0.02, # prior variable selection probability for standard Cox models
+  "a"      = -4, # hyperparameter in MRF prior
+  "b"      = 0.1, # hyperparameter in MRF prior
+  "G"      = simData$G, # hyperparameter in MRF prior
+  "lambda" = 3,
+  "nu0" = 0.05,
+  "nu1" = 5
+)
+
+# Run a 'Pooled' Bayesian Cox model with graphical learning
+
+set.seed(3346141)
+BayesSurvive_wrap <- function(use_cpp = FALSE) {
+  BayesSurvive(
+    survObj = list(dataset), model.type = "Pooled", MRF.G = FALSE,
+    hyperpar = hyperparPooled, initial = initial, nIter = 3,
+    verbose = TRUE, cpp = use_cpp
+  )
+}
+fit_R <- BayesSurvive_wrap(use_cpp = FALSE)
+fit_C <- BayesSurvive_wrap(use_cpp = TRUE)