R/quasi_symm.R

In Selbosh/scrooge: Statistical modelling of heterogeneous citation networks

#' Simulate a quasi-symmetric matrix
#'
#' @param n the dimension of the matrix
#' @param lambda the mean of the underlying Poisson random number generator
#' @param density the density (i.e. proportion of non-zeroes) of the symmetric component
#' @param dimnames names of the matrix dimensions, as a list
#'
#' @importFrom Matrix rsparsematrix drop0
#' @importFrom stats rnorm runif setNames
#'
#' @return A sparse, quasi-symmetric matrix of dimension `n` by `n`.
#'
#' @export
#'
#' @examples
#' rquasisymmetric(5)
rquasisymmetric <- function(n,
                            lambda = 10,
                            density = 1.00,
                            dimnames = list(cited = paste0('P', 1:n),
                                            citing = paste0('P', 1:n))) {
  S <- Matrix::rsparsematrix(n, n,
                             density = density,
                             symmetric = TRUE,
                             dimnames = dimnames,
                             rand.x = function(N) exp(rnorm(N)))
  a <- exp(runif(n))
  a <- setNames(a / sum(a), dimnames$cited)
  X <- Matrix::drop0(a * S)
  attr(X, 'S') <- S
  attr(X, 'a') <- a
  dimnames(X) <- dimnames
  X
}