Skip to contents

Warm-extends the fitted chain by further iterations and re-runs the convergence gate and the alignment on the grown draw set. The extension restores the sampler's saved random-number state, so the result is draw-for-draw identical to having run one longer chain, whatever happened in the session in between.

Usage

extend(fit, iterations = NULL, pivot = NULL, quiet = FALSE)

Arguments

fit

A bayesqm_fit created with keep_raw = TRUE.

iterations

Additional iterations; NULL (the default) doubles the chain. Must be a multiple of the fit's thin.

pivot, quiet

As in fit_bayesian().

Value

The extended bayesqm_fit.