2. Setting up calibration and priors outside the model file
Outside the model file, fixed parameters and priors are separated.
Fixed parameters can be given in a cell array:
params = {
'rho', 0.6
'beta', 0.998
...
'kappa', 161
};
or as a structure:
params = struct();
params.rho = 0.6;
params.beta = 0.998;
...
params.kappa = 161;
The parameterization is then pushed into the model object:
m = set(m, parameters = params);
Priors for Bayesian (and MLE) estimation are given as a structure; see Estimation for the five distinct parametrisations (uniform, mean+std, quantile, raw hyperparameters, user-defined PDF), the Dirichlet form for transition-matrix rows, and the full endogenous-priors facility.