Page 1 of 1

value of objective function in osr command

PostPosted: Wed Jun 15, 2011 12:20 pm
by nazanin
hello

i use osr command for finding optimal value of coefficients of taylor rule in dynare
dynare does iterations but dose not give any value for objective function
also give same value with initial value for optimal value of parameters

i don,t know how is it possible!?
and where is my mistake?

if you know please guide me
i attached my modfile


Thanks

Re: value of objective function in osr command

PostPosted: Thu Aug 25, 2011 3:30 pm
by SébastienVillemot
Your model has a unit root (you can see it with "check").

This means that the variance of some variables is infinite. This probably explains why "osr" does not converge.

Try to stationarize your model first.

Re: value of objective function in osr command

PostPosted: Thu Aug 25, 2011 5:23 pm
by olayenidynare
I had reported the error even before 4.2.1 version was released. I think it is a bug. The problem is that the m-file to report the objective value has been commented out. You may have to remove the comment yourself.

Re: value of objective function in osr command

PostPosted: Wed Sep 14, 2011 12:07 pm
by SébastienVillemot
olayenidynare wrote:I had reported the error even before 4.2.1 version was released. I think it is a bug. The problem is that the m-file to report the objective value has been commented out. You may have to remove the comment yourself.


I think you're confusing "osr" and "ramsey_policy".

There is indeed a problem with the objective value of "ramsey_policy" (thanks for reporting it), but the user who launched this thread talks about "osr".

Best,