run_model: Run the MCMC sampler for the Bayesian mixture model

In andburch/bayesianfingerprintr: Inferring the Age and Sex of Ancient Potters from Fingerprint Ridge Densities

Run the MCMC sampler for the Bayesian mixture model

Description

This function, as with all Bayesian models, will probably take a long time to run.

Usage

run_model(
  db,
  model.opt = c("2sex.2age", "1sex.1age", "2sex.1age", "1sex.2age", "compare1v2age",
    "compare1v2ageNEXT"),
  filename = "bayesmodel.txt",
  iters,
  thin,
  is.parallel = FALSE,
  DIC = FALSE,
  n.chains = 10,
  return.raw.samples = FALSE,
  n.adapt = 3000,
  n.burnin = 100,
  fix.biggest = FALSE,
  pseudopriors = NULL
)

Arguments

db	the data frame with observed MRB values (these must contained in a column named 'mrb')
model.opt	select which version of the model to run
filename	the name of the file that the JAGS model specifications will be saved to
iters	the number of iterations of the MCMC sampler
thin	thinning rate in the MCMC sampler. Must be a positive integer.
is.parallel	if parallel=TRUE, specify the number of CPU cores used. Defaults to total available cores or the number of chains, whichever is smaller.
DIC	option to report the deviance information criterion and the estimated number of parameters (pD). (perhaps useful for model selection)
n.chains	the number of chains in the MCMC sampler
return.raw.samples	whether the raw MCMC posterior samples be returned or a tidy database of samples linked to the actual input data
n.adapt	number of iterations to run in the JAGS adaptive phase. The default is NULL, which will result in the function running groups of 100 adaptation iterations (to a max of 10,000) until JAGS reports adaptation is sufficient. If you set n.adapt manually, 1000 is the recommended minimum value
n.burnin	number of iterations at the beginning of the chain to discard (i.e., the burn-in). Does not include the adaptive phase iterations.
fix.biggest	if TRUE, the data point with the largest MRB will be fixed as male. This can help if there are issues of model identifiability.
pseudopriors	(optional) this is for passing pseudopriors into the model when running a "super model"

Value

depending on the return.raw.samples parameter, either a MCMC list object or a data frame with the posterior distribution joined with the original data set. If the latter is chosen, the 'rowid.' column refers to the original row in the data frame that the estimates were made for.

Author(s)

Andrew Burchill, [e-mail]

Examples

## Not run: 
db = data.frame(mrb=c(rgamma(50,64,160),
                      rgamma(50,2500,5000)),
               location=sample(c("palace", "domestic"), 100, replace=TRUE)
               )
output = run_model(db, "2sex.2age", 15000, 5, F, F, 3, F)


## End(Not run)

andburch/bayesianfingerprintr documentation built on March 11, 2023, 8:53 p.m.