Estimation Problem


I’m running a model that’s suposed to make a bayesian estimation. It gives me a result, but this message appears:

(minus) the hessian matrix at the “mode” is not positive definite!
=> posterior variance of the estimated parameters are not positive.
You should try to change the initial values of the parameters using
the estimated_params_init block, or use another optimization routine.
Warning: The results below are most likely wrong!

What this means?

The message is pretty explicit. The mode you found is probably not a minimum as shown by the second order conditions (the Hessian matrix of second derivatives is not positive semi-definite). It might be that the message just derives from numerical inaccuracies, it could be that the mode-finder got stuck, or that you found a saddlepoint. Try mode_compute=6, which often helps.