Skip to contents

Generic method function for publication bias correction

Source: R/method.R
run_method.Rd

This function provides a unified interface to various publication bias correction methods. The specific method is determined by the first argument. See vignette("Adding_New_Methods", package = "PublicationBiasBenchmark") for details of extending the package with new methods

Usage

run_method(
  method_name,
  data,
  settings = NULL,
  silent = FALSE,
  fit_limit = NULL
)

Arguments

method_name

Character string specifying the method type

data

Data frame containing yi (effect sizes) and sei (standard errors)

settings

Either a character identifying a method version or list containing method-specific settings. An emty input will result in running the default (first implemented) version of the method.

silent

Logical indicating whether error messages from the method should be suppressed.

fit_limit

Optional numeric giving the maximum time, in minutes, that the method is allowed to run. NULL (the default) or a non-finite value imposes no limit. When the limit is exceeded, the fit is aborted and a failure result with convergence = FALSE and note = "time limit exceeded with <fit_limit> minutes" is returned. See the Time Limits section for the mechanism and its platform differences.

Value

A data frame with standardized method results

Time Limits

Methods that spend their time inside compiled sampling code (RoBMA via JAGS, RTMA and MMPH via Stan) do not return to R's evaluator and therefore cannot be stopped by R's own elapsed-time limit. fit_limit is consequently enforced by evaluating the method in a separate R process (callr::r_session) that is killed once the limit passes, which stops a fit regardless of what it is executing, on every platform.

That process is started on the first limited fit and reused by the following ones, adding roughly 0.05 seconds per fit; it is discarded and replaced whenever a fit is killed or the process dies. A fit that exceeds the limit therefore leaves no work behind, but also keeps nothing from the fits before it.

The worker has its own random number stream, which is seeded from the calling session for every fit. Results of methods that use randomness stay reproducible from the calling session's seed, but differ from those obtained without a limit.

Output Structure

The returned data frame follows a standardized schema that downstream functions rely on. All methods return the following columns:

  • method (character): The name of the method used.

  • estimate (numeric): The meta-analytic effect size estimate.

  • standard_error (numeric): Standard error of the estimate.

  • ci_lower (numeric): Lower bound of the 95% confidence interval.

  • ci_upper (numeric): Upper bound of the 95% confidence interval.

  • p_value (numeric): P-value for the estimate.

  • BF (numeric): Bayes Factor for the estimate.

  • convergence (logical): Whether the method converged successfully.

  • note (character): Additional notes describing convergence issues.

Some methods may include additional method-specific columns beyond these standard columns. Use get_method_extra_columns() to query which additional columns a particular method returns.

Examples

# Example usage with RMA method
data <- data.frame(
  yi = c(0.2, 0.3, 0.1, 0.4),
  sei = c(0.1, 0.15, 0.08, 0.12)
)
result <- run_method("RMA", data, "default")