rMultiNet: An R Package For Multilayer Networks Analysis

Overview

rMultiNet introduces an extension R package that includes a variety of traditional and state-of-the-art tensor decomposition methods for mixture multilayer networks analysis.The package is developed with the modular pipeline mode: generative modeling, embedding algorithms and visualization. rMultiNet aims to help study complex networks, especially mixture multilayer networks.

Getting Started

Prerequisites

Before using the rMultiNet, you need to install the following packages.

require(rTensor)
require(ggplot)
require(Matrix)
require(plotly)

Installation

Install the stable version of R-package from Git rMultiNet package directly with:

devtools::install_github("ChenyuzZZ73/rMultiNet")
bulid()
library(rMultiNet)

Usage example

The package is developed with the modular pipeline mode: generative modeling, embedding algorithms and visualization.

Generative modeling

library(rMultiNet)
GenerateMMSBM(n, m, L, K, d = NULL, r = NULL)
GenerateMMLSM(n, m, L, rank, U mean= 0.5, cmax =1, d, int type = ‘Uniform’, kernel fun = ‘logit’, scale par=1)

Also, rMultiNet provides three datasets for study：human malaria parasite gene network, worldwide food trading network and UN Commodity trading network.

load("~/Desktop/rMultiNet/data/malariagene/malaria.RData")

Emdedding algorithms

Take one of the algorithms for example.

InitializationMMSBM(tnsr, ranks=NULL)
PowerIteration(tnsr, ranks=NULL, type=”TWIST”, U 0 list, delta1=1000, delta2=1000, max iter = 25, tol = 1e-05)

Visulization

Embedding_network(network membership,L, paxis=2)
Community_cluster_km(embedding,type,cluster number)
Community_cluster_dbscan(embedding,type,eps value =.05, pts_value=5)

Citation

If you use rMultiNet or reference our tutorials in a presentation or publication, we would appreciate citations of our library.

Reference

• Bing-Yi Jing, Ting Li, Zhongyuan Lyu, and Dong Xia. Community detection on mixture multilayer networks via regularized tensor decomposition. The Annals of Statistics, 49 (6):3181–3205, 20
• Zhongyuan Lyu, Dong Xia, and Yuan Zhang. Latent space model for higher-order networks and generalized tensor decomposition. arXiv preprint arXiv:2106.16042, 2021.
• Xiaowen Dong, Pascal Frossard, Pierre Vandergheynst, and Nikolai Nefedov. Clustering with multi-layer graphs: A spectral perspective. IEEE Transactions on Signal Processing, 60(11):5820–5831, 2012