Slice4: Slice4

In Zink-Antoine/TLpack: Analyse TL measurement using Bayesian approach, MCMC, Gibbs and slice samplers

Slice4

Description

MC Analysis TL (slice +gibbs) following gibbs4.R Hickey (2006) Bayesian approach without expert judgment additional scalar variable T (Slice sampler)

Usage

Slice4(
  Dose,
  df.T,
  df.y,
  yn = 0,
  k = 1,
  n.iter,
  n.burnin = n.iter/2,
  n.thin = max(1, floor(n.iter - n.burnin)/10),
  threshold = min(df.y[df.y > 0])
)

Arguments

Dose	numeric (required) set of the irradiation doses
df.T	matrix (required) Luminescence data: the temperatures
df.y	matrix (required) Luminescence data; the luminescence signal
yn	numeric (with default) the reference luminescence signal. For MAAD TL, it is extrapolated to yn=0.
k	numeric (required) the number of parameters in the group involving sigma
n.iter	numeric (required) the number of iteration
n.burnin	numeric (with default) number of iterations to discard at the beginning (burn in). Default is n.iter/2, that is, discarding the first half of the simulations.
n.thin	numeric (with default) thinning rate. Must be a positive integer. Set n.thin > 1 to save memory and computation time if n.iter is large. Default is max(1, floor((n.iter-n.burnin) / 10)) which will only thin if there are at least 20 simulations.
threshold	numeric (with default) measured point below or equal to the threshold are rejected. Default is the minimum measured value.

Value

an (r × 5)-matrix,

column	Type	Description
alpha	numeric	intercept
beta	numeric	slope
sigma2	numeric	variance
T	numeric	Temperature
De	numeric	'True' Dose value

References

Gibbs sampler: Hickey, G. L. 2006. « The Linear Calibration Problem: A Bayesian Analysis ». PhD Thesis, PhD dissertation, University of Durham. 1–148. http://www.dur.ac.uk/g.l.hickey/dissertation.pdf.

chapter 6.3.6 and Appendix G.7

Slice sampler: Neal, R. 2003. « Slice Sampling ». Annals of Statistics 31 (3): 705‑67.

Examples

##load data
if(dev.cur()!=1) dev.off()
data(multiTL, envir = environment())
Dose<-multiTL$Dose
df.T<-multiTL$df.T
df.y<-multiTL$df.y
n.iter<-multiTL$n.iter
test<-Slice4(Dose=Dose,df.T=df.T,df.y=df.y,k=1,n.iter=n.iter)

Zink-Antoine/TLpack documentation built on April 14, 2025, 1:58 p.m.