Page 1 of 1

some problems in calibration

PostPosted: Sat Apr 15, 2017 2:06 pm
by zhanghuifd
Dear friends, sorry to post such some simple questions.
I encountered an error as below:
??? Error using ==> print_info at 74
Impossible to find the steady state. Either the model doesn't have a steady state, there are an infinity of steady states, or the guess values are too far from the solution

The residuals of some static equations are not equal to zero.

I also doubt that whether using "exp" or not for third-order approximation is correct, ie. exp(var).(and initial values of vars are in log).
In addition, I want to use the function of "norminv" in the model block to deduce instead of using normcdf, ie. Y = norminv(B). But it does not work. I also try with an external function but in vain.
Sorry for my incomplete questions.
Thanks a lot.

Re: some problems in calibration

PostPosted: Sat Apr 15, 2017 6:01 pm
by jpfeifer
Try not to do all at once. The first step is to get the model running at order=1 without any exp()-substitution. Once that works, doe the exp()-substitution, working from the running, unlogged version. Only then go to third order.
The norminv part will be tricky to achieve. Dynare does currently not support
Code: Select all
norminv

You can define it as an external function, but Dynare does not yet support this at order=3, see https://github.com/DynareTeam/dynare/issues/300

Re: some problems in calibration

PostPosted: Sun Apr 16, 2017 1:16 am
by zhanghuifd
Dear Prof. Pfeifer, Thanks for your advice!