Rank-normalized split R-hat, bulk and tail effective sample sizes, the
effective sample size of the mean and its Monte Carlo standard error:
the estimators of Vehtari et al. (2021), matching the posterior package.
Arguments
- draws
A matrix with one column per chain (or a list of
equal-length chains).
Value
A named numeric vector: rhat, ess_bulk, ess_tail,
ess_mean, mcse_mean.
Examples
set.seed(1)
mcmc_diagnostics(matrix(rnorm(4000), ncol = 4))
#> rhat ess_bulk ess_tail ess_mean mcse_mean
#> 1.000038e+00 3.941069e+03 4.133724e+03 3.941666e+03 1.649963e-02