plot_recovery: Plots to check parameter recovery

In qdercon/pstpipeline: Probabilistic Selection Task (PST) pipeline: parsing, analysis, simulation, and plotting

Plots to check parameter recovery

Description

plot_recovery produces correlation plots between input and output parameter values, plus a confusability matrix.

Usage

plot_recovery(
  raw_pars,
  rec_pars,
  test = FALSE,
  affect = FALSE,
  alpha_par_nms = NULL,
  plot_together = TRUE,
  incl_legend = !plot_together,
  pal = NULL,
  font = "",
  font_size = 11
)

Arguments

raw_pars	Parameters outputted from simulate_QL().
rec_pars	cmdstanr::summary() containing the fitted parameter values for the relevant model.
test	Boolean indicating whether recovered parameters are from the test phase.
affect	Boolean indicating whether recovered parameters are from an affect model.
alpha_par_nms	Option to rename learning rate parameters, defaults to the names from par_df.
plot_together	If TRUE, returns a panel with all plots plotted, otherwise a named list of plots is returned.
incl_legend	Boolean indicating whether to include legends for affect parameter plots (for the question types).
pal, font, font_size	Same as plot_import().

Value

A named list of ggplot objects.

A list or grid of ggplots.

Examples

## Not run: 
dir.create("outputs/cmdstan/simulated_data")

train_sim_2a <- simulate_QL(
  sample_size = 10,
  alpha_pos_dens = c(shape = 2, scale = 0.1), # default
  alpha_neg_dens = c(shape = 2, scale = 0.1), # default
  beta_dens = c(mean = 3, sd = 1) # default
)

mcmc_2a_train_sim <- fit_learning_model(
  train_sim_2a$sim, model = "2a", exp_part = "training",
  vb = FALSE, model_checks = FALSE, par_recovery = TRUE,
  outputs = c("raw_df", "stan_datalist", "summary", "draws_list"),
  out_dir = "outputs/cmdstan/simulated_data"
)

plot_recovery(train_sim_2a$pars, mcmc_2a_train_sim$summary)

## End(Not run)

qdercon/pstpipeline documentation built on June 1, 2025, 1:11 p.m.