predict.emc: Generate Posterior/Prior Predictives

In EMC2: Bayesian Hierarchical Analysis of Cognitive Models of Choice

Generate Posterior/Prior Predictives

Description

Simulate n_post data sets using the posterior/prior parameter estimates

Usage

## S3 method for class 'emc.prior'
predict(object, data = NULL, n_post = 50, n_cores = 1, n_trials = NULL, ...)

## S3 method for class 'emc'
predict(
  object,
  hyper = FALSE,
  n_post = 50,
  n_cores = 1,
  stat = c("random", "mean", "median")[1],
  ...
)

Arguments

object	An emc or emc.prior object from which to generate predictives
data	A data frame needed to exactly match the original design
n_post	Integer. Number of generated datasets
n_cores	Integer. Number of cores across which there should be parallellized
n_trials	An integer. If data isn't provided (although preferred), can generate data based on n_trials per cell of design
...	Optional additional arguments passed to get_pars or make_data
hyper	Boolean. Defaults to FALSE. If TRUE, simulates from the group-level (hyper) parameters instead of the subject-level parameters.
stat	Character. Can be mean, median or random (i.e., the default). Will take either random samples from the chain(s) or use the mean or median of the parameter estimates.

Value

A list of simulated data sets of length n_post

Examples

# based on an emc object ran by fit() we can generate posterior predictives
predict(samples_LNR, n_cores = 1, n_post = 10)

EMC2 documentation built on April 11, 2025, 5:50 p.m.