get_cov_gammas function

Covariance Matrices of Gamma Parameters

From a given MCMC, obtain a covariance matrix of the $gamma_ci$ ($i=1,2,3$) parameters for each country $c$. 1.1

get.cov.gammas(mcmc.set = NULL, sim.dir = NULL, burnin = 200, chain.id = 1)

Arguments

• mcmc.set: Object of class bayesTFR.mcmc.set. If it is NULL, the sim.dir is used to load existing simulation results.
• sim.dir: Directory with existing MCMC simulation results. It is only used if mcmc.set is NULL.
• burnin: Number of burn-in iterations to be discarded from the begining of the chain.
• chain.id: Identifier of the MCMC to be used. By default the first chain is used.

Details

In order to speed-up MCMC convergence, the package contains default values of gamma covariance that were obtained from a previous run (they can be loaded using data(proposal_cov_gammas_cii)). However, this function allows to obtain new values and overwrite the default values by passing the resulting object to the run.tfr.mcmc function as the proposal_cov_gammas argument.

Returns

A list with components: - values: An array of size nr_countries $x$ 3 $x$ 3 containing values of the covariance matrix of $gamma_ci$ ($i=1,2,3$) for each country $c$.

• country_codes: A vector of size nr_countries. A covariance matrix values[i,,] corresponds to a country with the code country_codes[i].

Author(s)

Leontine Alkema, Hana Sevcikova

See Also

run.tfr.mcmc

Examples

## Not run:

sim.dir <- file.path(find.package("bayesTFR"), "ex-data", "bayesTFR.output")
cov.gammas <- get.cov.gammas(sim.dir=sim.dir, burnin=0)
m <- run.tfr.mcmc(nr.chains=1, iter=10, proposal_cov_gammas=cov.gammas, verbose=TRUE)
## End(Not run)