We should add a method that computes the Fisher info at the best fit point using jax functionalities, so that we can compute it also for the cases in which generic likelihoods are present. This would also allow to simplify its calculation for quadratic fits and the definition of a generalised PCA.
We should add a method that computes the Fisher info at the best fit point using jax functionalities, so that we can compute it also for the cases in which generic likelihoods are present. This would also allow to simplify its calculation for quadratic fits and the definition of a generalised PCA.