Problem with estimation (Bayesian inference)


Im having trouble im my attempt to replicate an estimation (this is my first try ever!). I get this error message:

Log data density [Laplace approximation] is NaN.

Error using chol
Matrix must be positive definite with real diagonal.

Error in posterior_sampler_initialization (line 84)
d = chol(vv);

Error in posterior_sampler (line 60)
posterior_sampler_initialization(TargetFun, xparam1, vv,

Error in dynare_estimation_1 (line 465)

Error in dynare_estimation (line 105)

Error in curdia.driver (line 327)

Error in dynare (line 293)
evalin(‘base’,[fname ‘.driver’]) ;

I know its a problem in the MH optimization algoritm, but i have no idea how to solve it and what is the source of the problem.

Hope someone can help me.

Cúrdia et. al (2014).rar (163.6 KB)

I would recommend starting smaller.

  1. Where does your mode-file come from? You should run mode_compute=5 or similar.
  2. Are you sure your observation equations are correct? Your observable ffr has a very different mean than the one in the model.
  3. The parameter newphix is not identified.

Hey, thanks for the tips. My data was not treated properly. The parameter newphix was defined for other estimation (an alternative monetary policy rule). Now its working fine.