Endogenous prior option

Hi there,

Whenever I try to use the new endogenous prior option I get some problems with the positive definiteness of the the hessian matrix. I am using the mode_compute = 6 option. Any idea of why this happens ? If I remove this option I manage to go ahead with the estimation. I´m curious because I would like to see by how much the endogenous prior option changes my results, if any.


could you please provide an example.