Estimating NonLinear Neoclassical Growth

[mjs19]

I used to be able to run the following file, but it does not work anymore, giving me the error: Reference to non-existent field 'c'..
I saw a similar issue here in the forums where it was suggested to reinstall Dynare. I tried that but the error persisted. I have attached the file.
Any help is appreciated.

[jpfeifer]

The data file is missing. Which Dynare version are you using?

[mjs19]

Sorry about that, attached datafile here. I am using Dynare 4.4.3.
My actual objective is to find likelihood values for different parameters value, as in http://www.dynare.org/phpBB3/viewtopic.php?f=1&t=5547, but I couldnt even begin estimation...
Thanks.

[mjs19]

My observed data was incorrectly named, sorry about that. Anyways, in the other post I mentioned, in order to compute the likelihood,you suggest:

Code: Select all

global dataset_  bayestopt_       
[likelihood,junk1,junk2,exit_flag,ys,trend_coeff,info] = dsge_likelihood(xparam1,dataset_,options_,M_,estim_params_,bayestopt_,oo_);

In what order should xparam be defined? (which element of the vector correspond to each parameter?) Also, in your example, why do you need fs2000_mode? Also, I am getting positive values for likelihood, which is strange...
Thanks.

[jpfeifer]

What exactly is your question? The initial problem comes from the observed variable in the m-file not having the same name as the variable defined in the varobs command.