Dynare forums

Hi, I am trying to reproduce estimation results in Justiniano and Preston (2010).
However, I have failed to find posterior modes in estimation using dynare.

In fact, there are no particular contents in my code.
Estimation declaration is as follows.

Code: Select all

estimation(datafile=aus_data3,plot_priors=0,nograph,mode_compute=4,presample=4,prefilter=1,
diffuse_filter,mh_nblocks=1,mh_replic=50000,mh_jscale=0.10);

presample=4 is from the fact that 4 initial data points are excluded for initialization of Kalman filter.
prefilter=1 is because authors demeaned the series before the estimation.
Also diffuse_filter is added to deal with Blanchard - Kahn errors.
Others are basic options.

I think there is no mistakes in describing the model but continuously get the following error message.

Error using chol
Matrix must be positive definite.

Error in metropolis_hastings_initialization (line 68)
d = chol(vv);

Error in random_walk_metropolis_hastings (line 69)
[ ix2, ilogpo2, ModelName, MhDirectoryName, fblck, fline, npar, nblck, nruns, NewFile, MAX_nruns, d ] =
...

Error in dynare_estimation_1 (line 931)
feval(options_.posterior_sampling_method,objective_function,options_.proposal_distribution,xparam1,invhess,bounds,dataset_,options_,M_,estim_params_,bayestopt_,oo_);

Error in dynare_estimation (line 70)
dynare_estimation_1(var_list,dname);

Error in jp_2010 (line 331)
dynare_estimation(var_list_);

Error in dynare (line 120)
evalin('base',fname) ;


One thing strange is that prior declaration parts especially for shocks.

According to the paper,

To allow for a wide set of values a priori, we specify Inverse-Gamma I densities,
with infinite variance by fixing the degrees of freedom at 2.

That is why I wrote my code as

Code: Select all

stderr eps_ps,    0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_ys,    0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_is,    0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_a,     0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_m,     0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_g,     0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_rp,    0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;
    stderr eps_cp,    0.50,     1E-10,   10,    INV_GAMMA1_PDF,0.50,0.25;   

Is there any one can help me?

Thank you.

The data is missing.

Hi,
I tried running your model using mode_compute=6, and it runs without error. Am not an expert, but please try that out.

Regards,
Francis