smcUtils (version 0.2.2)

renormalize.weights: Renormalize weights

Description

This function creates a discrete probability distribution from a set of unnormalized (log) weights.

Usage

renormalize.weights(weights, log=FALSE, engine="R")

Arguments

weights

a vector of non-negative numbers

log

a logical indicating whether the weights are given as the logarithm of the weights

engine

run using "R" or "C" code

Value

• a vector of probabilities that sum to 1

See Also

resampling

Examples

ws = runif(10)
renormalize.weights(ws)

log.ws = dnorm(rnorm(10,1,0),log=TRUE)
renormalize.weights(log.ws,log=TRUE)