predict: Posterior Predictive Distribution

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

Description

Draw sample from the posterior predictive distribution.

Usage

predict(
  object,
  newdata = NULL,
  newdata_scale2 = NULL,
  newdata_scale3 = NULL,
  summary = TRUE,
  new_level = TRUE,
  cred = 0.95,
  digits = 3,
  ...
)

Arguments

object	An object of class blsmeta.
newdata	An optional data.frame for which to compute fit fitted values. Defaults to NULL, which then uses the original data used in blsmeta.
newdata_scale2	An optional data.frame for which to compute predictions for the level 2 variance component. Defaults to NULL, which then uses the original data used in blsmeta.
newdata_scale3	An optional data.frame for which to compute predictions for the level 3 variance component. Defaults to NULL, which then uses the original data used in blsmeta
summary	logical. Should the posterior samples be summarized (defaults to TRUE)?
new_level	logical. Should the predictive distribution be for new levels (defaults to TRUE)? See Details.
cred	numeric. credible interval (defaults to 0.95).
digits	numeric. The desired number of digits for the summarized estimates (defaults to 3).
...	Currently ignored.

Details

When new_level = TRUE, this provides the predictive distribution with respect to the moderators for the sub-models. For obtaining the predictive distribution for the observed levels, set this to FALSE which then includes the "random" effects.

Value

A data frame of fitted values (when summary = TRUE), or a matrix of the unsummarized posterior samples.

Examples

library(psymetadata)

fit <-  blsmeta(yi = yi, 
                vi = vi, 
                es_id = es_id,
                data = gnambs2020)
                
predict(fit)

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