BTSBM: Generates networks from binary tree stochastic block model

In HCD: Hierarchical Community Detection by Recursive Partitioning

Description

Generates networks from binary tree stochastic block model, with provided sequence of connection probability along the tree

Usage

BTSBM(n, d, a.seq, lambda, alpha = NULL, N = 1)

Arguments

n	number of nodes in the network
d	number of layers until leaves (excluding the root)
a.seq	the connection probability sequence along the tree, a_r, see details in the paper
lambda	average node degree, only used when alpha is not provided
alpha	the common scaling of the a_r sequence. So at the end, essentially the a_r sequence is a.seq*alpha
N	the number of networks to generate from the same model

Value

A list of objections of

A.list	the generated network adjacency matrices
B	the connection probability matrix between K communities, where K = 2^d
label	the vector of community labels for n nodes
P	the connection probability matrix between the n nodes. It is the expectation of adjacency matrices, except on the diagonal
comm.sim.mat	the binary string similarity matrix between communities
node.sim.mat	the binary string similarity matrix between nodes

Author(s)

Tianxi Li, Lihua Lei, Sharmodeep Bhattacharyya, Purnamrita Sarkar, Peter Bickel, and Elizaveta Levina.

Maintainer: Tianxi Li <[e-mail]>

References

Tianxi Li, Lihua Lei, Sharmodeep Bhattacharyya, Purnamrita Sarkar, Peter Bickel, and Elizaveta Levina. Hierarchical community detection by recursive partitioning. arXiv:1810.01509

Examples

dt <- BTSBM(n=1600,d=4,a.seq=0.2^seq(0,4),lambda=50)
A <- dt$A.list[[1]]

HCD documentation built on Aug. 24, 2019, 5:03 p.m.