Hello people,

When I runt the filtered_vars command in the estimation part (Dynare 4.1) and I choose 5000 replications, just to see if the command works, then everything is fine and I get the expected output.

But when i choose 100.000 replications then I get the following error :

Index exceeds matrix dimensions.

Error in DsgeSmoother (line 81)

constant = SteadyState(bayestopt_.mfys);

Error in prior_posterior_statistics (line 152)

[alphahat,etahat,epsilonhat,alphatilde,SteadyState,trend_coeff,aK] = …

Error in dynare_estimation_1 (line 1073)

prior_posterior_statistics(‘posterior’,data,gend,data_index,missing_value);

Error in dynare_estimation (line 62)

dynare_estimation_1(var_list,varargin{:});

Error in ctw_employer_surplus (line 2190)

dynare_estimation(var_list_);

Error in dynare (line 132)

evalin(‘base’,fname) ;

Does anyone have a clue why it might not work with more replications?

Thank you!

Could you try Dynare 4.3.3?

Unfortunately I have to use Dynare 4.1.

If it’s a bug with this version I’ll just let it be…

Could you then provide me with the mod-file (e.g. private message). I would be interested to find out whether this is a bug that might be still present.

Sorry for my late reply jpfeifer,

I’ve checked the forum for similar issues (4.1 version seem to have a problem) and I think I’ll first give it a try with version 4.3.2 and version 4.3.3 (using matlab R2012b) and see how it goes.

Anyway, how can I create the simulated series for variables i.e one-sided predicted values (such as the attached file of Adolfson’s et al.(2005)) using Dynare? This is my big question. I would be thankful for any advise. Does this have anything to do with the filtered_vars command?

Best regards.

Data and one-sided predicted value.pdf (48.4 KB)

One-sided predicted values should be the filtered variables. Note that they are not really simulated variables.