runMCMCGlobal: Run MCMC estimation for hierarichal models with multiple...

In MJAlexander/distortr: Temporal smoothing methods for demographic time series

Description

Run MCMC estimation of time series data for multiple countries using JAGS.

Usage

runMCMCGlobal(method, input.data, order = NULL, matern.cov = TRUE,
  cs.arma = NULL, cs.smoothing = TRUE, time.trend = FALSE,
  nserror.estimated = TRUE, nchains = 3, nburnin = 1000, niter = 2000,
  nthin = 1, model.file.path = NULL, model.save.file.path = "R/model.txt")

Arguments

method	The method of smoothing to implement (choices: ar, arma, splines, gp)
input.data	List of required input data. See processData, getSplinesData and getGPData to get required data in compatible form.
order	The order of splines penalization (either 1 or 2)
matern.cov	Whether or not to use Matern covariance function. Default is TRUE.
cs.arma	whether ARMA parameter(s) are country specific. If 'FALSE', parameter is global.
cs.smoothing	whether smoothing paramter is country specific. If 'FALSE', smoothing parameter is global.
time.trend	if 'TRUE' a linear time trend is estimated.
nserror.estimated	whether to estimate non-sampling error. IF 'FALSE', fixed sampling error is inputted.
nchains	Number of MCMC chains
nburnin	Number of iterations to throw away as burn in.
niter	Number of total iterations.
nthin	Degree of thinning of MCMC chains
model.file.path	Text file which contains the model to be fitted. If NULL, the text file is drawn from the models folder.
model.save.file.path	Path to save model, if written.

Value

A JAGS model object

See Also

processData, getSplinesData, getGPData

Examples

nyears <- 100
prop.sample <- 0.7
obs.err <- TRUE
sigma.y <- 0.5
seed <- 123
method <- 'splines'
params <- list(sigma.alpha = 1, order = 1)
df <- simulateFluctuations(nyears, prop.sample, method, params, obs.err, sigma.y)
df$se <- 1
mod <- runMCMC(input.data = df, nyears = 100, method = "splines", order = 1,
nchains = 4, nburnin = 100, niter = 100+3000, nthin = 3)

MJAlexander/distortr documentation built on July 17, 2020, 4:06 p.m.