eval_prior_list: Evaluates Several Prior Distributions

In alexanderrobitzsch/mdmb: Model Based Treatment of Missing Data

Evaluates Several Prior Distributions

Description

The function eval_prior_list evaluates several prior distributions specified in a list. The function eval_prior_list_sumlog computes the sum of the logarithms of all prior values.

Usage

eval_prior_list(par, par_prior, log=FALSE, eps=1e-50)

eval_prior_list_sumlog(par, par_prior, use_grad=1)

Arguments

par	Parameter vector
par_prior	List of prior distributions specified in a list (see Examples)
log	Logical indicating whether the logarithm of the prior should be computed
eps	Decimal to be added to the prior to avoid computation of the logarithm for values of zero
use_grad	Integer value for computation value for gradient

Value

Vector or a numeric value

See Also

sirt::prior_model_parse

Examples

#############################################################################
# EXAMPLE 1: Evaluation of prior values
#############################################################################

# normal distribution
b0 <- list( "dnorm", list(mean=0,sd=100) )
# t distribution with one degree of freedom (Cauchy distribution)
b1 <- list( "dt", list(df=1) )
# define list of priors
beta_prior <- list( b0, b1 )
# parameter value
beta <- c( 0.3, 1 )

#-- evaluate priors
mdmb::eval_prior_list(par=beta, par_prior=beta_prior)
mdmb::eval_prior_list_sumlog(par=beta, par_prior=beta_prior)

alexanderrobitzsch/mdmb documentation built on July 18, 2024, 11:14 p.m.