I am trying to find an optimal monetary policy using Bayesian estimation.
Stochastic simulations and historical decomposition are working fine but I am getting a following error message: Log data density [Laplace approximation] is NaN.
My data file is detrended. Can anyone help me to understand any issues? I am happy to provide codes and datafile if needed.
Thanks for your help.
What do the
mode_check-plots tell you?
i v also encountered the same problem.
the model was fine before estimation(mode_compute=0)
but when i trying to do the estimation there comes error(mode_compute=4):
Log data density [Laplace approximation] is NaN.
errorcut.fig (93.2 KB)
Mode finding was clearly not successful in getting to the mode, but rather got stuck in the increasing part. The main issue seems to be
SE_eg running into the upper bound, causing a corner solution.