Rsampletrees (version 0.1)

setWeights: Set weights for the sampletrees proposal distributions

Description

Set the probabilities/weights for sampling which proposal distribution to apply in a given step of the MCMC chain.

Usage

setWeights(args, WeightFile = NULL, weightmat = NULL)

Arguments

args

An object of class `pars' that stores arguments for a sampletrees run

WeightFile

The name of file to write out the proposal distribution weights. If missing it is set to "weights"

weightmat

A matrix with the sampling weights for the proposal distributions. The first column contains the sampling probability for the proposal distribution listed in the second column. If missing, default values are used. Defaults are given in the Details section.

Value

args

An object of class `pars' with the tag WeightFile set to the file specified.

The weights are also written out to the file specified by WeightFile.

Details

This function is used to write a set of weights to a file and to set the file name argument for WeightFile in the sampletrees settings object. The user can use default weights (if weightmat is NULL) or they can specify alternative weights by providing an argument to weightmat.

The matrix weightmat must be a 2 column matrix of maximum 7 rows. The first column contains the sampling probability for the proposal distribution listed in the second column. The first column must sum to 1. The second column will consist only of some subset of the numbers 1 through 7.

The default values for the sampling weights are:

haplotype data: (0.1,0.1,0.4,0.2,0.2) for update types 1-5

genotype data: (0.1,0.1,0.35,0.1,0.1,0.125,0.125) for update types 1-7

References

TBD

Examples

runpars=newArgs()
runpars=setWeights(runpars, "weights-h.txt")

runpars=newArgs(DataType="g")
runpars=setWeights(runpars, "weights-g-2.txt", 
weightmat=cbind(c(0.05,0.05,0.4,0.2,0.15,0.15),c(1:4,6:7)))