rase (version 0.2-2)

post.mcmc: Post-MCMC Handling

Description

Takes the result from an mcmc run (e.g., output from rase) and applies a burnin and a thinning. Optionally it can be transformed into a ggs object that the ggmcmc package can use (for mcmc visualization).

Usage

post.mcmc(res, burnin = 1000, thin = 10, as.ggmcmc = TRUE)

Arguments

res

The output from any mcmc run in (e.g., output from rase). It is effectively a matrix where each column is a different parameter and each row is one iteration of the mcmc sampler.

burnin

Number of iterations to disregard as burnin.

thin

Sample every ith iteration. If no iterations are to be discarted, it should be thin = 1. Note that if thin = 0, then no iteration is saved at all.

as.ggmcmc

If TRUE (default), the object will be transformed into a ggs object.

Value

• If as.ggmcmc = TRUE (the default), the function returns a ggs object. This can be used as input to the ggs_* functions in the ggmcmc package. If as.ggmcmc = FALSE, the function returns a matrix where each column is a different parameter and each row is one iteration of the mcmc sampler.

See Also

rase, bm_ase, ggs

Examples

#attach the data
data(rase_data, package = 'rase')
	
# Explore the data
str(mcmc)

# discard the first 200 iterations as burnin 
#and keep every 10th iteration.
gg_dat <- post.mcmc(mcmc, burnin=200, thin = 10, as.ggmcmc = TRUE)

# density using ggmcmc for dispersal rates (sigma).
require(ggmcmc)
ggs_density(gg_dat, family = 'sigma')