Convert multi-chain bkmrfit to mcmc.list for coda MCMC diagnostics
Converts a kmrfit.list (from the bkmrhat package) into
an mcmc.list object from the coda package.The
coda package enables many different types of MCMC diagnostics,
including geweke.diag, traceplot and
effectiveSize. Posterior summarization is also available,
such as HPDinterval and summary.mcmc.
Using multiple chains is necessary for certain MCMC diagnostics, such as
gelman.diag, and gelman.plot.
## S3 method for class 'list.bkmrfit.list' as.mcmc(x, ...)
x |
object of type kmrfit.list (from bkmrhat package) |
... |
arguments to |
An mcmc.list object
# following example from https://jenfb.github.io/bkmr/overview.html set.seed(111) library(coda) dat <- bkmr::SimData(n = 50, M = 4) y <- dat$y Z <- dat$Z X <- dat$X set.seed(111) Sys.setenv(R_FUTURE_SUPPORTSMULTICORE_UNSTABLE="quiet") future::plan(strategy = future::multiprocess, workers=2) # run 2 parallel Markov chains (more usually better) fitkm.list <- kmbayes_parallel(nchains=2, y = y, Z = Z, X = X, iter = 1000, verbose = FALSE, varsel = FALSE) mcmcobj = as.mcmc.list(fitkm.list) summary(mcmcobj) # Gelman/Rubin diagnostics won't work on certain objects, # like delta parameters (when using variable selection), # so the rstan version of this will work better (does not give errors) try(gelman.diag(mcmcobj)) # lots of functions in the coda package to use plot(mcmcobj) # both of these will also fail with delta functions (when using variable selection) try(gelman.plot(mcmcobj)) try(geweke.plot(mcmcobj)) closeAllConnections()
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.