Pareto-smoothed importance sampling leave-one-out cross-validation

This is observation-level LOO: it measures within-study interpolation (leaving out one IPD patient or one reconstructed pseudo-observation), which is not the scientific question in a disconnected network. It does not validate cross-gap prediction (a new study, a held-out treatment contrast, or a held-out sub-network); a good pointwise LOO can coexist with a wrong cross-gap extrapolation. Grouped leave-one-study-out is not yet implemented, so unit accepts only "observation".

Usage

# S3 method for class 'cpaic_mlnmr'
loo(x, unit = "observation", ...)

Arguments

x

A cmlnmr() fit.

unit

Predictive unit; only "observation" is supported.

...

Passed to loo::loo.matrix().

Value

A psis_loo object from the loo package.