Classify malaria infections using Bayesian MCMC

Runs the Bayesian MCMC engine across all sites to classify patient samples as recrudescence or reinfection. Called automatically by MalReBay, but can also be used directly for a step-by-step workflow.

Usage

classify_infections(
  imported_data,
  mcmc_config = system.file("extdata", "default_mcmc_config.rds", package = "MalReBay"),
  n_workers = 1,
  verbose = TRUE,
  suppress_warnings = TRUE
)

Arguments

imported_data

A list returned by import_data.

mcmc_config

Path to an MCMC configuration Excel file, or a named list of parameters. Defaults to the bundled configuration.

n_workers

Number of parallel workers. Defaults to 1.

verbose

Logical. Print progress messages. Defaults to TRUE.

suppress_warnings

Logical. If TRUE (default), silences the MCMC sampler's own console warnings about divergent transitions, treedepth, and E-BFMI (the "N of M transitions ended with a divergence" messages cmdstanr prints as soon as sampling finishes, independently of verbose). Set to FALSE to see them – e.g. while tuning adapt_delta in mcmc_config for a problematic site. This only affects what gets printed; the same diagnostics are always available afterwards via summarise_results's convergence table.

Value

A named list of raw MCMC results per site, or NULL if no valid results are produced. Pass this to summarise_results.

See also

import_data, summarise_results, MalReBay

Examples

if (FALSE) { # \dontrun{
imported <- import_data()
results  <- classify_infections(imported)
} # }