dCRP: Compute Probability Mass of a Partition Under the Two...

In sams: Merge-Split Samplers for Conjugate Bayesian Nonparametric Models

Compute Probability Mass of a Partition Under the Two Parameter Chinese Restaurant Process (CRP)

Description

Compute Probability Mass of a Partition Under the Two Parameter Chinese Restaurant Process (CRP)

Usage

dCRP(partition, mass = 1, discount = 0, log = FALSE)

Arguments

partition	A numeric vector of cluster labels, or a matrix whose rows are numeric vectors of cluster labels
mass	A numeric value indicating the mass parameter in the CRP, which must be greater than the -discount argument
discount	A numeric value on the interval [0,1), indicating the discount parameter of the two parameter CRP
log	A logical value indicating whether results should be returned on the log scale

Value

A numeric vector of probabilities, or log probabilities if log = TRUE.

Examples

partitions <- matrix(c(0,0,0,0,0,
                       0,0,0,0,1,
                       0,0,0,1,2,
                       0,0,1,2,3,
                       0,1,2,3,4), ncol = 5, nrow = 5, byrow = TRUE)

# discount = 0 shows higher probability for lower quantity of components
dCRP(partitions, mass = 1, discount = 0, log = FALSE)

# discount = 0.5 shows higher probability for higher quantity of components
dCRP(partitions, mass = 1, discount = 0.5, log = FALSE)

sams documentation built on April 20, 2022, 1:06 a.m.