ppsbm (version 1.0.0)

generateDynppsbm: Dynppsbm data generator

Description

Generates data under the Dynamic Poisson Process Stochastic Blockmodel (dynppsbm).

Usage

generateDynppsbm(intens, Time, n, prop.groups, directed = TRUE)

Value

Simulated data, latent group variables and intensities \(\alpha^{(q,l)}\).

Arguments

intens

List of intensity functions \(\alpha^{(q,l)}\), each one is a list of 2 components:

• intens : a positive function. The intensity function \(\alpha^{(q,l)}\)

• max : positive real number. An upper bound on \(\alpha^{(q,l)}\)

Time

Positive real number. [0,Time] is the total time interval of observation.

n

Total number of nodes, \(1\le i \le n\).

prop.groups

Vector of group proportions (probability to belong to a group), should be of length \(Q\)

directed

Boolean for directed (TRUE) or undirected (FALSE) case.

If directed then intens should be of length \(Q^2\), else of length \(Q*(Q+1)/2\).

References

MATIAS, C., REBAFKA, T. & VILLERS, F. (2018). A semiparametric extension of the stochastic block model for longitudinal networks. Biometrika. 105(3): 665-680.

Examples

# Generate data from an undirected graph with n=10 individuals and Q=2 clusters

# equal cluster proportions
prop.groups <- c(0.5,0.5)

# 3 different intensity functions:
intens <- list(NULL)
intens[[1]] <- list(intens= function(x) 100*x*exp(-8*x),max=5)
    # (q,l) = (1,1)
intens[[2]] <- list(intens= function(x) exp(3*x)*(sin(6*pi*x-pi/2)+1)/2,max=13)
    # (q,l) = (1,2)
intens[[3]] <- list(intens= function(x) 8.1*(exp(-6*abs(x-1/2))-.049),max=8)
    # (q,l) = (2,2)

# generate data:
obs <- generateDynppsbm(intens,Time=1,n=10,prop.groups,directed=FALSE)

# latent variables (true clustering of the individuals)
obs$z

# number of time events:
length(obs$data$time.seq)

# number of interactions between each pair of individuals:
table(obs$data$type.seq)