Skip to contents

Upgrade a saved fit object to the current BayesRTMB class definitions

Source: R/upgrade_fit.R
upgrade_fit.Rd

`upgrade_fit()` rebuilds an existing fit object with the currently loaded BayesRTMB R6 class definitions. This is useful for saved fit objects created by older package versions when newer methods have been added to the fit classes.

Usage

upgrade_fit(fit, upgrade_model = TRUE)

Arguments

fit

A BayesRTMB fit object. Supported classes are MCMC, VB, MAP, and classic fit objects.

upgrade_model

Logical; if `TRUE`, rebuild the embedded model object from its stored data, parameter list, and model functions. If rebuilding fails, the original model is kept and a warning is issued.

Value

A fit object of the same type, using the currently loaded class definitions.

Details

The function does not refit the model or recompute estimates. It only copies the stored estimation results into a new fit object. When `upgrade_model` is `TRUE`, the embedded `RTMB_Model` object is also rebuilt from the model components stored in the fit, so newer model methods are available as well.