sample_prior: Draw Samples from the Prior Distributions

In donaldRwilliams/blsmeta: Bayesian Location-Scale Meta-Analysis

Description

This function is useful for visualizing the prior distributions.

Usage

sample_prior(object, iter = 5000, ...)

Arguments

object	An object of class make_prior.
iter	numeric. The number of posterior samples per chain (defaults to 5000, excluding warmup).
...	Currently ignored.

Details

Note that parameters include beta (location or effect size), gamma (level-two scale model), and eta (level-three scale model). Importantly, the scale model priors are on the log-scale (use exp).

Value

A data frame including the posterior samples.

Examples

library(psymetadata)

prior <- c(assign_prior(param = "(Intercept)", 
           prior = "dnorm(0, 1)", dpar = "location"),
           assign_prior(param = "(Intercept)", 
           prior = "dnorm(-2, 1)", 
           dpar = "scale", level = "two")
           )

priors <- make_prior(yi = yi, 
                     vi = vi, 
                     prior = prior,
                     es_id = es_id,
                     study_id = study_id,
                     data = gnambs2020)

samps <- sample_prior(priors, iter = 50000)
hist(samps$beta)

donaldRwilliams/blsmeta documentation built on Dec. 20, 2021, 12:12 a.m.