R/toAdjMatrix.R

In baycn: Bayesian Inference for Causal Networks

# toAdjMatrix
#
# Creates an adjacency matrix from the DNA of the current graph. The DNA
# is converted to an adjacency matrix to determine the parents of each node
# in order to calculate its log likelihood.
#
# @param coordinates A matrix with the row and column coordinates of the
# edges from the adjacency matrix from the MRPC function. The row numbers of
# each nonzero element make up the first row in the matrix and the column
# numbers of each nonzero element make up the second row of the matrix.
#
# @param graph A vector containing the directions of the edges in the
# graph and the log likelihood of the graph.
#
# @param nEdges The number of edges in the graph.
#
# @param nNodes The number of nodes in the graph.
#
# @return An adjacency matrix given the current edge states.
#
toAdjMatrix <- function (coordinates,
                         graph,
                         nEdges,
                         nNodes) {

  # Create a pxp matrix of 0s to be converted to an adjacency matrix.
  adjMatrix <- matrix(0,
                      nrow = nNodes,
                      ncol = nNodes)

  # create the adjacency matrix with the edges oriented according to the DNA.
  # This will be used in determining which function to use when calculating the
  # log likelihood for each node.
  for(e in 1:nEdges) {

    if (graph[[e]] == 0) {

      adjMatrix[coordinates[1, e], coordinates[2, e]] <- 1

    } else if (graph[[e]] == 1) {

      adjMatrix[coordinates[2, e], coordinates[1, e]] <- 1

    }

  }

  return(adjMatrix)

}