swap_chains {LDATS}

Conduct a set of among-chain swaps for the ptMCMC algorithm

Description

This function handles the among-chain swapping based on temperatures and likelihood differentials.

This function was designed to work within TS and specifically est_changepoints. It is still hardcoded to do so, but has the capacity to be generalized to work with any estimation via ptMCMC with additional coding work.

Usage

swap_chains(chainsin, inputs, ids)

Arguments

chainsin

Chain configuration to be evaluated for swapping.

inputs

Class ptMCMC_inputs list, containing the static inputs for use within the ptMCMC algorithm.

ids

The vector of integer chain ids.

Details

The ptMCMC algorithm couples the chains (which are taking their own walks on the distribution surface) through "swaps", where neighboring chains exchange configurations (Geyer 1991, Falcioni and Deem 1999) following the Metropolis criterion (Metropolis et al. 1953). This allows them to share information and search the surface in combination (Earl and Deem 2005).

Value

list of updated change points, log-likelihoods, and chain ids, as well as a vector of acceptance indicators for each swap.

References

Earl, D. J. and M. W. Deem. 2005. Parallel tempering: theory, applications, and new perspectives. Physical Chemistry Chemical Physics 7: 3910-3916. link.

Falcioni, M. and M. W. Deem. 1999. A biased Monte Carlo scheme for zeolite structure solution. Journal of Chemical Physics 110: 1754-1766. link.

Geyer, C. J. 1991. Markov Chain Monte Carlo maximum likelihood. In Computing Science and Statistics: Proceedings of the 23rd Symposium on the Interface. pp 156-163. American Statistical Association, New York, USA. link.

Metropolis, N., A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller. 1953. Equations of state calculations by fast computing machines. Journal of Chemical Physics 21: 1087-1092. link.

Examples

  data(rodents)
  document_term_table <- rodents$document_term_table
  document_covariate_table <- rodents$document_covariate_table
  LDA_models <- LDA_set(document_term_table, topics = 2)[[1]]
  data <- document_covariate_table
  data$gamma <- LDA_models@gamma
  weights <- document_weights(document_term_table)
  data <- data[order(data[,"newmoon"]), ]
  saves <- prep_saves(1, TS_control())
  inputs <- prep_ptMCMC_inputs(data, gamma ~ 1, 1, "newmoon", weights, 
                               TS_control())
  cpts <- prep_cpts(data, gamma ~ 1, 1, "newmoon", weights, TS_control())
  ids <- prep_ids(TS_control())
  for(i in 1:TS_control()$nit){
    steps <- step_chains(i, cpts, inputs)
    swaps <- swap_chains(steps, inputs, ids)
    saves <- update_saves(i, saves, steps, swaps)
    cpts <- update_cpts(cpts, swaps)
    ids <- update_ids(ids, swaps)
  }