Mposterior (version 0.1.2)

Mposterior-package: Robust and Scalable Bayes via a Median of Subset Posterior Measures.

Description

Implementation of Weiszfeld algorithm for estimating M-posterior for robust and scalable Bayesian inference (see Minsker et al., 2014).

Details

Package:

Mposterior

Type:

Package

Version:

0.1.1

Date:

2014-05-31

License:

GPL (>= 3)

LazyLoad:

yes

findWeiszfeldMedian is the workhorse function that estimates M-posterior given samples from the subset posteriors using the Weiszfeld algorithm of Minsker et al. (2014). M-posterior is the median of subset posteriors in the space of probability measures.

References

Minsker, S., Srivastava, S., Lin, L., and Dunson, D.B. (2014). Robust and Scalable Bayes via a Median of Subset Posterior Measures. http://arxiv.org/abs/1403.2660

See Also

findWeiszfeldMedian

Examples

set.seed(12345)
## list that contains subset posterior samples from 2-dim Gaussian density
subAtomList <- vector("list", 5)
subAtomList[[1]] <- cbind(rnorm(100, mean = 1),  rnorm(100, mean = 1))
subAtomList[[2]] <- cbind(rnorm(100, mean = -1),  rnorm(100, mean= -1))
subAtomList[[3]] <- cbind(rnorm(100, mean = -1),  rnorm(100, mean = 1))
subAtomList[[4]] <- cbind(rnorm(100, mean = 1),  rnorm(100, mean = -1))
subAtomList[[5]] <- cbind(rnorm(100, mean = 2),  rnorm(100, mean = 2))
library(Mposterior)
medPosterior <- findWeiszfeldMedian(subAtomList, sigma = 0.1, maxit = 100, tol = 1e-10)
medPosterior
summary(medPosterior)
plot(medPosterior)