Build a standard diagnostics table from MCMC samples

Constructs a basic diagnostics table from an mcmc.list, including:

• target: parameter name,

• ESS: minimum effective sample size across chains,

• AE_ESS_per_it: algorithmic efficiency (ESS / iterations),

• ESS_per_sec: computational efficiency (ESS / runtime),

• time_s_per_ESS: seconds per effective sample,

• Rhat: Gelman-Rubin convergence diagnostic,

• Family: top-level family (prefix before the first [).

Usage

compute_diag_from_mcmc(samples, runtime_s)

Arguments

samples

An mcmc.list (or a single mcmc / matrix).

runtime_s

Numeric scalar; wall-clock runtime in seconds.

Value

A data.frame named diag_tbl.