Generic (expected) log-predictive density (deprecated)

As of loo 3.0.0, elpd() is deprecated. Please use measure_elpd() instead. For full predictive performance workflows, see insample_pred_measure() and loo_pred_measure(). See vignette("migration-guide", package = "loo") for a full mapping table.

Usage

elpd(x, ...)

# S3 method for class 'array'
elpd(x, ...)

# S3 method for class 'matrix'
elpd(x, ...)

Arguments

x

A log-likelihood array or matrix. The Methods (by class) section, below, has detailed descriptions of how to specify the inputs for each method.

...

Currently ignored.

Details

The return type differs: elpd() returns class "elpd_generic" with elpd and ic in pointwise; measure_elpd() returns class "measure".

The elpd() methods for arrays and matrices can compute the expected log pointwise predictive density for a new dataset or the log pointwise predictive density of the observed data (an overestimate of the elpd). The elpd() function is an S3 generic and methods are provided for 3-D pointwise log-likelihood arrays and matrices.

Methods (by class)

• elpd(array): An \(I\) by \(C\) by \(N\) array, where \(I\) is the number of MCMC iterations per chain, \(C\) is the number of chains, and \(N\) is the number of data points.

• elpd(matrix): An \(S\) by \(N\) matrix, where \(S\) is the size of the posterior sample (with all chains merged) and \(N\) is the number of data points.

See also

measure_elpd(), insample_pred_measure(), loo_pred_measure(), and the vignette Holdout validation and K-fold cross-validation.

Examples

# \dontrun{
# Deprecated:
LLarr <- example_loglik_array()
elpd(LLarr)
#> Warning: 'elpd.array' is deprecated.
#> Use 'measure_elpd' instead.
#> See help("Deprecated")
#> 
#> Computed from 1000 posterior draws and 32 log-likelihood terms using the generic elpd function.
#> 
#>      Estimate  SE
#> elpd    -80.3 3.2
#> ic      160.5 6.5
# ->
measure_elpd(LLarr)
#> 
#> Computed from 1000 draws by 32 observations.
#> 
#>      Estimate    SE
#> elpd    -80.3   3.2
# }