WAIC is a more fully Bayesian approach for estimating the out-of-sample expectation based on the log pointwise posterior predictive density
WAIC(mod, bsim = NA, nsim = 100)an object of class lm, glm or mer
an object of class simMer (optional), if provided computing time is reduced.
number of simulations used to describe the posterior distributions, if bsim is provided, this number is taken from bsim.
log pointwise posterior predictive density: the logarithms of the predictive density integrated over the posterior distribution of the model parameters summed over all observations.
an estimate for the number of effective parameters
a second estimate for the number of effective parameters
WAIC based on pwaic1
WAIC based on pwaic2
We implemented the formulas given in Gelman et al. (2014) p 173. We hope that the implementation is correct! For hierarchical (mixed) models, the function gives the WAIC that measures predictive fit for the groups in the data (not for new groups). For hierarchical models the predictive fit could be measured for each level of the data. But this flexibility is not yet implemented in the WAIC function.
Gelman, A., Carlin, J.B., Stern, H.S., Dunson, D.B., Vehtari, A. & Rubin, D.B. (2014) Bayesian Data Analysis, Third edn. CRC Press.
Watanabe, S. (2010) Applicable Information Criterion in Singular Learning Theory. Journal of Machine Learning Research, 11, 3571-3594.
# NOT RUN {
data(pondfrog1)
mod1 <- glm(frog ~ ph + waterdepth + temp, data=pondfrog1, family=poisson)
mod2 <- glm(frog ~ + waterdepth + temp, data=pondfrog1, family=poisson)
mod3 <- glm(frog ~ ph + + temp, data=pondfrog1, family=poisson)
mod4 <- glm(frog ~ ph + waterdepth , data=pondfrog1, family=poisson)
WAIC(mod1)
WAIC(mod2)
WAIC(mod3)
WAIC(mod4)
# }
Run the code above in your browser using DataLab