BchronDensity {Bchron}

Non-parametric phase model

Description

This function runs a non-parametric phase model on 14C and non-14C ages via Gaussian Mixture density estimation

Usage

BchronDensity(
  ages,
  ageSds,
  calCurves,
  pathToCalCurves = system.file("data", package = "Bchron"),
  dfs = rep(100, length(ages)),
  numMix = 50,
  iterations = 10000,
  burn = 2000,
  thin = 8,
  updateAges = FALSE,
  store_density = TRUE
)

Arguments

ages

A vector of ages (most likely 14C)

ageSds

A vector of 1-sigma values for the ages given above

calCurves

A vector of values containing either intcal20, shcal20, marine20, or normal (older calibration curves such as intcal13 are also supported). Should be the same length the number of ages supplied. Non-standard calibration curves can be used provided they are supplied in the same format as those previously mentioned and are placed in the same directory. Normal indicates a normally-distributed (non-14C) age.

pathToCalCurves

File path to where the calibration curves are located. Defaults to the system directory where the 3 standard calibration curves are stored

dfs

Degrees-of-freedom values for the t-distribution associated with the calibration calculation. A large value indicates Gaussian distributions assumed for the 14C ages

numMix

The number of mixture components in the phase model. Might need to be increased if the data set is large and the phase behaviour is very complex

iterations

The number of iterations to run for

burn

The number of starting iterations to discard

thin

The step size of iterations to keep

updateAges

Whether or not to update ages as part of the MCMC run. Default is FALSE. Changing this to TRUE will improve performance but will fit a slightly invalid model

store_density

Whether or not to store the density and age grid. Useful for plotting the output in other packages

Details

This model places a Gaussian mixture prior distribution on the calibrated ages and so estimates the density of the overall set of radiocarbon ages. It is designed to be a probabilistic version of the Oxcal SUM command which takes calibrated ages and sums the probability distributions with the aim of estimating activity through age as a proxy.

Value

An object of class BchronDensityRun with the following elements:

• thetaThe posterior samples of the restricted ages

• pPosterior samples of the mixture proportions

• muValues of the means of each Gaussian mixture

• calAgesThe calibrated ages from BchronCalibrate

• GThe number of mixture components. Equal to numMix

• age_gridA grid of ages used for the final density estimate

• densityThe density estimate based on the above age grid

See Also

Bchronology, BchronRSL, BchronDensityFast for a faster approximate version of this function

Examples

# Read in some data from Sluggan Moss
data(Sluggan)

# Run the model
SlugDens <- with(
  Sluggan,
  BchronDensity(
    ages = ages,
    ageSds = ageSds,
    calCurves = calCurves
  )
)

# plot it
plot(SlugDens)