mcmcSampler: pMCMCsampler

In aaronm70/batMods: fitsSEIR

Description

pMCMCsampler

Usage

mcmcSampler(
  initParams,
  randInit = F,
  proposer = sequential.proposer,
  sdProps,
  maxSddProps,
  niter = 100,
  particleNum = 100,
  nburn = 100,
  monitoring = 2,
  adaptiveMCMC = T,
  proposerType = "seq",
  startAdapt = 1000,
  adptBurn = 200,
  acceptanceRate = 0.25,
  tell = 5,
  cluster = F,
  oDat,
  stoch = T,
  priorFunc = NA,
  switch = 2500,
  switchBlock = 2500,
  likelihoodFunc = likelihoodFunc1,
  juvenileInfection = F
)

Arguments

randInit	T then randomly sample initial parameters instead of initParams value
proposer	proposal function, multivariate block (adaptiveMCMC must = T) or sequential can have adaptive or not adaptive tuning
sdProps	standard deviation for proposal distributions - in adaptive this is the starting sd
maxSddProps	maximum values for sd proposals if using adaptive MCMC
niter	number of iterations to run the MCMC for
nburn	number of mcmc iterations to burn
monitoring	0 = no monitoring, > 0 prints more progress information
adaptiveMCMC	T/F if true uses tuning for s.d. of parmater proposal distributions based on acceptance ratios
proposerType	"seq" or "block", seq for sequential proposing, block for blocked proposing, blocks based on var-covar matrix - block proposing only available when using adaptiveMCMC
startAdapt	starting iteration for adapting
acceptanceRate	acceptance rates, for adaptive sequential use string, one rate for each param, for block use single value
tell	print monitoring information every x number of iterations
cluster	T/F if true, using dide cluster, if false running locally
initial	parameter guess
particles	number of particles for particle filter
adaptBurn	burn n number of iterations for defining var-covar matrix in block proposing

Value

mcmc

aaronm70/batMods documentation built on Sept. 8, 2021, 7:05 a.m.