Create priors for BayesianTools

Helper function for creating log-priors compatible with BayesianTools package

pda.create.btprior(prior.sel)

Arguments

prior.sel

data.frame containing prior distributions of the selected parameters

Value

out Prior class object for BayesianTools package

Details

prior.sel must contain the following columns:

• distn – String describing a distribution; e.g. norm for dnorm, rnorm, etc.

• parama, paramb – First and second parameters, respectively, of the corresponding distribution

Optionally, prior.sel may also contain the following columns:

• param_name – Parameter name, which will be carried through to the prior object and sampler

• lower, upper – Lower and upper bounds, respectively. These can be leveraged by the BayesianTools samplers.

• best – Best guess for a parameter estimate. BayesianTools can also use this, though I'm not sure how...

Author

Istem Fer, Alexey Shiklomanov