pwaic: Compute the effective number of parameters 'pwaic' for a...

Description Usage Arguments Value

View source: R/waic.R

Description

For each observation, compute the variance of the log likelihood. Note that we are computing the variance of the log likelihood. The total pwaic for the model is the sum of these, but we return the individual values for use in computing standard errors.

Usage

1
pwaic(pointwise_ll)

Arguments

pointwise_ll

List returned from pointwise_loglike.

Value

A vector of pwaic values averaged over Monte Carlo samples. Each observation will have one corresponding point in the vector.


JGCRI/gcamland documentation built on Oct. 6, 2020, 5:30 p.m.