bayesTFR (version 4.0-0)

run.tfr3.mcmc: Running Markov Chain Monte Carlo for Parameters of Total Fertility Rate in Phase III

Description

Runs (or continues running) MCMCs for simulating phase III total fertility rate, using a Bayesian hierarchical version of an AR(1) model.

Usage

run.tfr3.mcmc(sim.dir, nr.chains = 3, iter = 50000, thin = 10, 
    replace.output = FALSE, my.tfr.file = NULL, buffer.size = 100, 
    use.extra.countries = FALSE, 
    mu.prior.range = c(0, 2.1), rho.prior.range = c(0, 1 - .Machine$double.xmin), 
    sigma.mu.prior.range = c(1e-05, 0.318), sigma.rho.prior.range = c(1e-05, 0.289), 
    sigma.eps.prior.range = c(1e-05, 0.5), 
    mu.ini = NULL, mu.ini.range = mu.prior.range, 
    rho.ini = NULL, rho.ini.range = rho.prior.range, 
    sigma.mu.ini = NULL, sigma.mu.ini.range = sigma.mu.prior.range, 
    sigma.rho.ini = NULL, sigma.rho.ini.range = sigma.rho.prior.range, 
    sigma.eps.ini = NULL, sigma.eps.ini.range = sigma.eps.prior.range, 
    seed = NULL, parallel = FALSE, nr.nodes = nr.chains, compression.type = "None", 
    auto.conf = list(max.loops = 5, iter = 50000, iter.incr = 20000, nr.chains = 3, 
                    thin = 60, burnin = 10000), 
    verbose = FALSE, verbose.iter = 1000, ...)
        
continue.tfr3.mcmc(sim.dir, iter, chain.ids=NULL, 
    parallel = FALSE, nr.nodes = NULL, auto.conf = NULL,
    verbose=FALSE, verbose.iter = 1000, ...)

Value

An object of class bayesTFR.mcmc.set which is a list with two components:metaAn object of class bayesTFR.mcmc.meta.mcmc.listA list of objects of class bayesTFR.mcmc, one for each MCMC.

Details

The MCMCs are stored in sim.dir in a subdirectory called phaseIII. It has exactly the same structure as phase II MCMCs described in run.tfr.mcmc.

References

Raftery, A.E., Alkema, L. and Gerland, P. (2013). Bayesian Population Projections for the United Nations. Statistical Science, in press.

See Also

run.tfr.mcmc, get.tfr3.mcmc

Examples

sim.dir <- tempfile()
# Runs Phase II MCMCs (must be run before Phase III)
m <- run.tfr.mcmc(nr.chains=1, iter=5, output.dir=sim.dir, verbose=TRUE)
# Runs Phase III MCMCs
m3 <- run.tfr3.mcmc(sim.dir=sim.dir, nr.chains=2, iter=50, thin=1, verbose=TRUE)
m3 <- continue.tfr3.mcmc(sim.dir=sim.dir, iter=10, verbose=TRUE)
summary(m3, burnin=10)
unlink(sim.dir, recursive=TRUE)