spikeSlabGAM
provides a formula-based interface for specifying models
and takes care of (most of) the housekeeping. Sampling of
the chains is done in parallel using package
multicore
if available. If not, a socket cluster
set up with snow
is used where available. Use
options(cores=foo)
to set the (maximal) number of
processes spawned by the parallelization. If
options()$cores
is unspecified, snow
uses
8.spikeAndSlab(y, X,
family = c("gaussian", "binomial", "poisson"),
hyperparameters = list(), model = list(),
mcmc = list(), start = list())