select_TS: Select the best Time Series model

In LDATS: Latent Dirichlet Allocation Coupled with Time Series Analyses

Select the best Time Series model

Description

Select the best model of interest from an TS_on_LDA object generated by TS_on_LDA, based on a set of user-provided functions. The functions default to choosing the model with the lowest AIC value.

Presently, the set of functions should result in a singular selected model. If multiple models are chosen via the selection, only the first is returned.

Usage

select_TS(TS_models, control = list())

Arguments

TS_models	An object of class TS_on_LDA produced by TS_on_LDA.
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.

Value

A reduced version of TS_models that only includes the selected TS model. The returned object is a single TS model object of class TS_fit.

Examples

  data(rodents)
  document_term_table <- rodents$document_term_table
  document_covariate_table <- rodents$document_covariate_table
  LDAs <- LDA_set(document_term_table, topics = 2:3, nseeds = 2)
  LDA_models <- select_LDA(LDAs)
  weights <- document_weights(document_term_table)
  formulas <- c(~ 1, ~ newmoon)
  mods <- TS_on_LDA(LDA_models, document_covariate_table, formulas,
                    nchangepoints = 0:1, timename = "newmoon", weights)
  select_TS(mods)

LDATS documentation built on Sept. 19, 2023, 5:08 p.m.