Calculate massively parallelised tied diffusion scores — rs_tied_diffusion

Helper function to calculate in parallel on the same (unweighted) network the tied diffusions as fast as possible. Can be used for permutation.

Usage

rs_tied_diffusion_parallel(
  node_names,
  from,
  to,
  weights,
  diffusion_scores_1,
  diffusion_scores_2,
  summarisation_fun,
  undirected
)

Arguments

node_names: String vector. Name of the graph nodes.
from: String vector. The names of the from edges from the edge list.
to: String vector. The names of the to edges from the edge list.
weights: Optional weight vector. If NULL, defaults to 1.0 as weight for all edges.
diffusion_scores_1: List. The first set of personalised vectors for the page rank reset values. Each element must sum to 1 and be of same length of node_names!
diffusion_scores_2: List. The second set of personalised vectors for the page rank reset values. Each element must sum to 1 and be of same length of node_names!
summarisation_fun: String. One of c("min", "max", "avg"). Which type of summarisation function to use to calculate the tied diffusion.
undirected: Boolean. Is this an undirected graph.

Value

A matrix of the scores with each row representing a tied diffusion of of diffusion_scores_1 and diffusion_scores_2 lists (in order), and each column representing the value of the tied diffusion for this node.