Learn R Programming

optismixture (version 0.1)

Optimal Mixture Weights in Multiple Importance Sampling

Description

Code for optimal mixture weights in importance sampling. Workhorse functions penoptpersp() and penoptpersp.alpha.only() minimize estimated variances with and without control variates respectively. It can be used in adaptive mixture importance sampling, for example, function batch.estimation() does two stages, a pilot estimate of mixing alpha and a following importance sampling.

Copy Link

Version

Install

install.packages('optismixture')

Monthly Downloads

7

Version

0.1

License

GPL-2

Maintainer

Hera He

Last Published

August 25th, 2015

Functions in optismixture (0.1)

Do plain monte carlo with target density

Internal function. Get the row index in the stacked sample matrices for the $b^{th}$ batch

Internal function. convert mixture proportions to mixture sample size with a fixed total sample size

batch.estimation

Two stage estimation, a pilot estimate of mixing alpha and a following importance sampling, with or without control variates

Internal function. With stratified samples, calculate the variance of the estimate from importance sampling without control variates

penalized optimization of the constrained linearized perspective function

compatible.test

Test the compatibility of user defined functions fname, rpname, rqname, dpname, dqname with mixture.param

Optimal Mixture Weights in Multiple Importance Sampling

mixture.is.estimation

For a given mixture weight alpha, use importance sample with or withour control variates for estimation

penoptpersp.alpha.only

penalized optimization of the constrained linearized perspective function

get.initial.alpha

Internal function. Calculate the initial alpha vector for the optimization of alpha with a lower bound constraint

do.mixture.sample

Internal function. sample from the mixture distribution $q_{\alpha}$