ppsbm (version 1.0.0)

kernelIntensities: Direct kernel estimator intensities

Description

Compute smooth intensities with direct kernel estimation of intensities relying on a classification \(\tau\). This can be used with the values \(\tau\) obtained on a dataset with mainVEM function.

Usage

kernelIntensities(
  data,
  tau,
  Q,
  n,
  directed,
  rho = 1,
  sparse = FALSE,
  nb.points = 1000 * data$Time
)

Arguments

data

List with 3 components:

• time.seq - Vector of observed time points of the events (length \(M\)).

• type.seq - Vector of observed types of node pairs (as encoded through convertNodePair of the events (length \(M\))).

• Time - [0,Time] is the total time interval of observation.

tau

Matrix with size \(Q\times n\) and values in \((0,1)\), containing the (estimated) probability that cluster \(q\) contains node \(i\).

Q

Total number of groups.

n

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

directed

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

rho

Either 1 (non sparse case) or vector with length \(Q(Q+1)/2\) (undirected case) or \(Q^2\) (directed case) with (estimated) values for the sparsity parameters \(\rho^{(q,l)}\). See Section S6 in the supplementary material paper of Matias et al. (Biometrika, 2018) for more details.

sparse

Boolean for sparse (TRUE) or not sparse (FALSE) case.

nb.points

Number of points for the kernel estimation.

Details

Warning: sparse case not implemented !!!

Examples

# The generated_sol_kernel solution was generated calling mainVEM
# with kernel method on the generated_Q3$data dataset.
# (50 individuals and 3 clusters)

data <- generated_Q3$data

n <- 50
Q <- 3


# Compute smooth intensity estimators
sol.kernel.intensities <- kernelIntensities(data,generated_sol_kernel$tau,Q,n,directed=FALSE)