The manual mentions Laplace Approximation probably back in 2023 or maybe earlier. I can’t find it in the rstan
interface.
The best I found is the optimizing
function, that returns the MAP solution with an option to get the hessian, which I believe can be used to derive the covariance as partially outlined in the above page.
At least for the parameters. For the transformed parameters, I would need to perform the transformation myself.
Am I missing something, or is there anything more provided for the Laplace Approximation?