Page 1 of 1

Warning: Log of zero.

PostPosted: Tue Feb 17, 2015 8:49 pm
by ekaya
ı'm working on SW 2007 model. I run model an MH algorithm run. It's okey. but in command window, there is massage below. This warnings comtinue, but I couldn't see he posterior mean and the confidence interval results. What can I do now ? Please help me

Warning: Log of zero.
> In marginal_density at 84
In dynare_estimation_1 at 889
In dynare_estimation at 62
In usmodel at 402
In dynare at 120
Warning: Log of zero.
> In marginal_density at 84
In dynare_estimation_1 at 889
In dynare_estimation at 62
In usmodel at 402
In dynare at 120
Warning: Log of zero.
> In marginal_density at 84
In dynare_estimation_1 at 889
In dynare_estimation at 62
In usmodel at 402
In dynare at 120
Warning: Log of zero.
> In marginal_density at 84
In dynare_estimation_1 at 889
In dynare_estimation at 62
In usmodel at 402
In dynare at 120
Warning: Log of zero.
> In marginal_density at 84
In dynare_estimation_1 at 889
In dynare_estimation at 62
In usmodel at 402
In dynare at 120
Warning: Log of zero.
> In marginal_density at 84
In dynare_estimation_1 at 889
In dynare_estimation at 62
In usmodel at 402
In dynare at 120

Re: Warning: Log of zero.

PostPosted: Wed Feb 18, 2015 2:25 pm
by jpfeifer
There must be something wrong in the estimation step before.

Re: Warning: Log of zero.

PostPosted: Wed Feb 18, 2015 2:50 pm
by ekaya
Mr. Johannes Pfeifer,
Again thank you very much for your answer. I couldn't find which step is wrong, Can you help me, please?

Re: Warning: Log of zero.

PostPosted: Thu Feb 19, 2015 9:07 am
by jpfeifer
You need to post the mod-file and the data.

Re: Warning: Log of zero.

PostPosted: Thu Feb 19, 2015 9:45 am
by ekaya
Dear Mr. Johannes Pfeifer,
I sended you my files from private
Thanks

Re: Warning: Log of zero.

PostPosted: Mon Feb 23, 2015 9:02 am
by ekaya
Dear Mr. Johannes Pfeifer,
I'm waitinig your answer because of my estimatin problem in Dynare
I studied again in my log of zero problem, but It couldn't be solved.
Thanks

Re: Warning: Log of zero.

PostPosted: Tue Feb 24, 2015 9:26 am
by jpfeifer
Your data file is inconsistent with the observation equations. You need to construct data that is consistent with the way the Smets-Wouters model was set up. See their paper and Pfeifer(2013): "A Guide to Specifying Observation Equations for the Estimation of DSGE Models" https://sites.google.com/site/pfeiferecon/Pfeifer_2013_Observation_Equations.pdf.

Re: Warning: Log of zero.

PostPosted: Tue Feb 24, 2015 12:59 pm
by ekaya
Thanks again.
I solved my problem. The program now works.