Adjusted R-squared and LOO-CV

[Babu Lawrence]

Dear INLA community,

I used INLA to fit three spatiotemporal models and compared model performance using WAIC and DIC. The reviewer is interested in adjusted R-squared values to understand the amount of variance explained by the model. Is it possible to determine the adjusted R-squared value in inla, or is there an alternative to explain the model variation in inla models?

Also, how does leave-one-out cross-validation work in INLA? I have computed CPO in my models, which is the leave-one-out predictive check in INLA models.

Thank you very much for your help.

Lawrence