Create priors from an existing distribution model

Often it can make sense to fit an additional model to get a grasp on the range of values that "beta" parameters can take. This function takes an existing BiodiversityDistribution object and creates PriorList object from them. The resulting object can be used to add for instance priors to a new model.

Usage

get_priors(mod, target_engine, ...)

# S4 method for class 'ANY,character'
get_priors(mod, target_engine, ...)

Arguments

mod

A fitted DistributionModel object. If instead a BiodiversityDistribution object is passed to this function, it simply returns the contained priors used for estimation (if any).

target_engine

A character for which the priors should be created.

...

Other parameters passed down.

Value

A PriorList object, or NULL when no compatible priors can be created.

Note

Not all engines support priors in similar ways. See the vignettes and help pages on that topic!

See also

Other prior: BARTPrior(), BARTPriors(), BREGPrior(), BREGPriors(), GDBPrior(), GDBPriors(), GLMNETPrior(), GLMNETPriors(), INLAPrior(), INLAPriors(), STANPrior(), STANPriors(), XGBInteractionPrior(), XGBInteractionPriors(), XGBPrior(), XGBPriors(), add_priors(), priors(), rm_priors()

Examples

if (FALSE) { # \dontrun{
 mod <- distribution(background) |>
    add_predictors(covariates) |>
    add_biodiversity_poipo(points) |>
    engine_inlabru() |>
    train()
 get_priors(mod, target_engine = "BART")
} # }