hmodelBUGS: Bayesian general linear regression analysis of current trial...

In shanRpackage/pprincrt: Bayesian Power Prior Design and Analysis of Cluster Randomized Trials

Description

hmodelBUGS returns posterior samples of treatment effect

Usage

hmodelBUGS(hData, hForm, file.dir, file.name, file.rm = F, OpenBUGS.dir,
  OpenBUGS.seed = 1, family, nchain, niter, nburnin, nthin)

Arguments

hData	historical trial data. It should be in a data frame.
hForm	model for historical trial data. It should be a formula object.
file.dir	directory to save model BUGS file and input and output files of OpenBUGS.
file.name	the name of the model BUGS file.
file.rm	whether to remove all files after Bayesian computations are done.
OpenBUGS.dir	directory to install OpenBUGS package.
OpenBUGS.seed	the random number generator state of OpenBUGS.
family	outcome distribution.
nchain	number of MCMC chains.
niter	length of MCMC chain.
nburnin	burnin period of MCMC chain.
nthin	thinning rate of MCMC chain.

Value

The posterior samples of treatment effect

Examples

## Not run: 
#' my historical data;
my.hData <- data.frame(y=rbinom(3,50,0.5),n=rep(50,3),x=c(1,2,3))

my.hsample <- hmodelBUGS(hData=my.hData,hForm=y|n ~ x,
file.dir=tempdir(),file.name='hmodelfile',file.rm=T, OpenBUGS.dir=NULL,
OpenBUGS.seed=14, family="binomial",nchain=2,niter=5000,nburnin=2500,nthin=5)

## End(Not run)

shanRpackage/pprincrt documentation built on May 23, 2019, 1:11 p.m.