py-block-imh

The block IMH algorithm is an improvement of the Independent Metropolis-Hastings (IMH) algorithm. It reduces the variance of estimates based on the IMH output, by making use of all the proposed values instead of making only use of the accepted values, and by considering permutations of the proposals. Additionally it can strongly benefit from parallel computational power. Reference: Using parallel computation to improve Independent Metropolis--Hastings based estimation, by Pierre Jacob (Universite Paris-Dauphine and CREST, France), Christian P. Robert (Universite Paris-Dauphine and CREST, France), Murray H. Smith (NIWA, Wellington, New Zealand). The article is available here: http://arxiv.org/abs/1010.1595

py-block-imh
IMHprobit.py
autoLaunchToy.py
IMH.py
gpl.txt
comparisonIS.R
.hgignore
prof
results
raoblackwell.py
plotAll.py
localfolder.py
probitplots.R
comparisonPermutations.R
launchProbit.py
pat.py
comparisonRB.R
preliminaries.R
probit.py
pickletoRdata.py
launchAll.py
launchToy.py
.hg
autoLaunchProbit.py
probitplots2.R
permutations.py
README.txt