consensusNet: consensusNet

In INetTool: Integration Network

consensusNet

Description

This function computes the INet Algorithm for the construction of a **Consensus Network**.

Usage

consensusNet(
  adjL,
  threshold = 0.5,
  tolerance = 0.1,
  theta = 0.04,
  nitermax = 50,
  ncores = 2,
  verbose = TRUE
)

Arguments

adjL	list of weighted adjacency matrix with weights in [0,1]. Same name in rows and columns for all the matrices.
threshold	threshold for the construction of the Consensus (default 0.5). Used in the last step on the similar graphs.
tolerance	the tolerance of differences between similar graphs for the construction of the Consensus (default 0.1).
theta	importance to give to the neighbourhood part of the weight (default 0.04).
nitermax	maximum number of iteration before stopping the algorithm (default 50).
ncores	number of CPU cores to use (default is 2). We suggest to use ncores equal to the number of graphs to integrate.
verbose	flag for verbose output (default as TRUE).

Value

a list of 3 types: $graphConsensus the Consensus Network, $Comparison the Jaccard weighted distances between the graphs calculated in each iteration, $similarGraphs the similar graphs before the Thresholding

Examples

data("adjL_data")
consensusNet(adjL_data)

INetTool documentation built on March 20, 2026, 1:07 a.m.