FitConv: Fit and Converge a list of Jags models

In MarcoDVisser/FDPtools: Optimized tools for working with forest dynamics plot data

Description

FitConv - runs a Jags model, or list of models, and checks for convergence using gelman stats. It then returns a list of MCMC samples, a jags model and DIC values.

Usage

FitConv(JAGSdata = NULL, modellist = NULL, gelmanthr = 1.05,
  StepIt = 1000, MaxIt = 10000, FinalSampleLength = 500, hlist = NULL,
  ...)

Arguments

JAGSdata	A pre-prepared jags data object
modellist	A vector containing the location and names of the jags model files. These files should be read in R, as a modelstring. These files must also generate a vector of model parameter names to be returned (in a object called interestlist)
gelmanthr	Gelman threshold values to except fit
StepIt	what is the standard number of MCMC samples to take before checking for convergence?
MaxIt	What is the maximum number of MCMC samples
FinalSampleLength	How many posterior sample do you need when the model is returned?
hlist	Use this when you have a list of higher parameters that should converge first before convergence of lower parameters (in interestlist) should be checked.
...	additional parameters to be passed to jags.model

Author(s)

Marco D. Visser

MarcoDVisser/FDPtools documentation built on May 7, 2019, 2:49 p.m.