Removed n_cores option for updateTheta since it is not used anymore

NAMESPACE

@@ -25,7 +25,6 @@ export(set_attributes_to_edges_visnetwork)
 export(strip_cvn)
 export(updateTheta)
 export(updateY)
-export(updateZ)
 export(updateZ_wrapper)
 export(visnetwork)
 export(visnetwork_cvn)

R/update-Theta.R

@@ -9,12 +9,11 @@
 #' @param Sigma A list with empirical covariance matrices \eqn{{\Sigma}}
 #' @param n_obs A \eqn{m}-dimensional vector with the number of observations per graph 
 #' @param rho The \eqn{\rho} ADMM's penalty parameter (Default: 1) 
-#' @param n_cores Number of cores used (Default: 1)
 #' 
 #' @return A list with matrices with the new values of \eqn{\Theta}
 #'        
 #' @export
-updateTheta <- function(m, Z, Y, Sigma, n_obs, rho = 1, n_cores = 1) { 
+updateTheta <- function(m, Z, Y, Sigma, n_obs, rho = 1) { 
 
   # we use the fact that we can separate the optimization problem for each 
   # individual graph i = 1, 2,..., m 
@@ -32,20 +31,4 @@ updateTheta <- function(m, Z, Y, Sigma, n_obs, rho = 1, n_cores = 1) {
     # The new Theta matrix
     Q %*% diag(Lambda) %*% t(Q)
   })
-
-  # mclapply(1:m, function(i) {
-  #   
-  #   # perform an eigendecomposition 
-  #   eigen_decomposition <- eigen(Sigma[[i]] - rho * Z[[i]]/ n_obs[i] + rho * Y[[i]] / n_obs[i])
-  #   
-  #   # obtain matrices Q and Lambda (Lambda is a diagonal matrix, so first only the diagonal is stored)
-  #   Q <- eigen_decomposition$vectors 
-  #   Lambda <- eigen_decomposition$values
-  # 
-  #   # Lambda is updated 
-  #   Lambda <- n_obs[i]/(2 * rho) * (-Lambda + sqrt(Lambda^2 + 4*rho/n_obs[i]))
-  #   
-  #   # The new Theta matrix
-  #   Q %*% diag(Lambda) %*% t(Q)}, mc.cores = n_cores # number of cores used
-  # )
 }

sandbox/profile-package.R

@@ -45,7 +45,7 @@ W <- CVNSim::create_weight_matrix("full", m = m)
 library(tictoc)
 tic()
 
-cvn_new <- CVN(data, W, n_cores = 4, maxiter = 1000, lambda1 = lambda1, lambda2 = lambda2, verbose = F)
+cvn_new <- CVN(data, W, n_cores = 1, maxiter = 1000, lambda1 = lambda1, lambda2 = lambda2, verbose = F)
 toc()