Prepare and update the data structures to save the ptMCMC output

prep_saves creates the data structure used to save the output from each iteration of the ptMCMC algorithm, which is added via update_saves. Once the ptMCMC is complete, the saved data objects are then processed (burn-in iterations are dropped and the remaining iterations are thinned) via process_saves.

This set of functions was designed to work within TS and specifically est_changepoints. They are still hardcoded to do so, but have the capacity to be generalized to work with any estimation via ptMCMC with additional coding work.

prep_saves(nchangepoints, control = list())

update_saves(i, saves, steps, swaps)

process_saves(saves, control = list())

Arguments

nchangepoints	integer corresponding to the number of change points to include in the model. 0 is a valid input (corresponding to no change points, so a singular time series model), and the current implementation can reasonably include up to 6 change points. The number of change points is used to dictate the segmentation of the data for each continuous model and each LDA model.
control	A list of parameters to control the fitting of the Time Series model including the parallel tempering Markov Chain Monte Carlo (ptMCMC) controls. Values not input assume defaults set by TS_control.
i	integer iteration index.
saves	The existing list of saved data objects.
steps	Chain configuration after within-temperature steps.
swaps	Chain configuration after among-temperature swaps.

Value

list of ptMCMC objects: change points ($cpts), log-likelihoods ($lls), chain ids ($ids), step acceptances ($step_accepts), and swap acceptances ($swap_accepts).

Examples

# \donttest{
  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)
  }
  process_saves(saves, TS_control())
# }