ssMousetrack (version 1.1.7)

check_prior: Check prior distributions

Description

Check prior distributions

Usage

check_prior(priors = NULL)

Value

a matrix contatining priors (numeric codes, see Details) and their parameters

Arguments

priors

(list) a list of arguments specifying priors for each parameter involved in the model (see Details). If priors=NULL then predefined priors will be returned.

Details

The function is used to specify the prior-related arguments of the state-space modeling function run_ssm. Priors are specified in terms of distributions and associated parameters as implemented in the rstan package. The available options are as follows:

• lognormal(mu,sigma) (code = 1, required parameters = 2)

• normal(mu,sigma) (code = 2, required parameters = 2)

• normal(mu,sigma)T(min,max) (code = 201, required parameters = 4)

• chi_square(df) (code = 3, required parameters = 1)

• inv_chi_square(df) (code = 4, required parameters = 1)

• gamma(alpha,beta) (code = 5, required parameters = 2)

• pareto(min,alpha) (code = 6, required parameters = 2)

• uniform(min,max) (code = 7, required parameters = 2)

This is an internal function, generally not to be called by the user.

Examples

if (FALSE) {
## Define priors for all the paramaters 
priors_list <- list("lognormal(1,1)","normal(2,3)T(0,10)","normal(3,1)")
priors_out <- check_prior(priors_list)
print(priors_out)

## Define priors for some of the paramaters 
priors_list <- list(NULL,"pareto(1,1.2)",NULL)
priors_out <- check_prior(priors_list)
print(priors_out)

## Use pre-defined vague priors for all the parameters
priors_list <- list(NULL,NULL,NULL)
priors_out <- check_prior(priors_list)
print(priors_out)
}