The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
The goal of mcunit is to provide unit tests for MCMC and Monte Carlo methods. It extends the package testthat.
You can install the current version of mcunit from bitbucket with:
::install_bitbucket("agandy/mcunit") devtools
This shows how to test if a sampler has a specific mean:
<- function(n) rnorm(n,mean=3.2,1)
sampler expect_mc_iid_mean(sampler,mean=3.2)
expect_mc_iid_mean(sampler,mean=3.5)
#> Error: Test failed with p-value=2.734643e-20 in iteration 1
This shows check of a simple MCMC sampler
<- list(genprior=function() rnorm(1),
object gendata=function(theta) rnorm(5,theta),
stepMCMC=function(theta,data,thinning){
<- function(x) prod(dnorm(data,x))*dnorm(x)
f for (i in 1:thinning){
= rnorm(1,mean=theta,sd=1)
thetanew if (runif(1)<f(thetanew)/f(theta))
<- thetanew
theta
}
theta
}
)expect_mcmc_reversible(object)
## And now with an error in the sampler:
## sampling until sample is accepted.
$stepMCMC <- function(theta,data,thinning){
object<- function(x) prod(dnorm(data,x))*dnorm(x)
f for (i in 1:thinning){
repeat{
= rnorm(1,mean=theta,sd=1)
thetanew if (runif(1)<f(thetanew)/f(theta)) break;
}<- thetanew
theta
}
theta
}expect_mcmc_reversible(object,control=list(n=1e4))
#> Error: Test failed with p-value=3.586124e-11 in iteration 1
These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.