Estimation, matrix for SVD function

Identification/Sensitivity
1

Dear all,

I have a problem doing Bayesian estimation of my model.
The error message says:

Error using svd
Input to SVD must not contain NaN or Inf.
Error in rank (line 14)
s = svd(A);
Error in identification_analysis (line 232)
if iflag || rank(MIM)>rank(tildaM)
Error in dynare_identification (line 308)
identification_analysis(params,indx,indexo,options_ident,dataset_, dataset_info, prior_exist, name_tex,1,parameters,bounds);
Error in code (line 2044)
dynare_identification(options_ident);
Error in dynare (line 235)
evalin(‘base’,fname) ;
Error in run (line 305)

From this I understand that there is something wrong with this matrix A.
So, I have a few questions. What is this matrix exactly? How is it calculated? And, most importantly, why an inf or Nan may appear in this matrix?

I highly appreciate your help

Regards,
Alex

2

Which Dynare version are you using? Most probably, I would need to see the files.

3

It’s the version 4.5.7

Can it be a problem due to this Dynare version?

Thank you for your reply, Prof. Pfeifer!

4

Yes, it could be. That’s why I would need to see the files.

5

Dear Professor Pfeifer,

I have the same issue when dealing with the identification of the model using Dynare version 4.5.7
Currently, I cannot find the answer to this problem. Could you please investigate this and tell me what I should do to solve this?
I really appreciate your help and thank you so much!

I would attach the files here for your reference.

paper2_steadystate.m (2.9 KB)
paper2.mod (4.8 KB)

Best regards,

Anh Pham

6

Yes, this was a problem in Dynare 4.5. Please use this steady state file to run 4.6.4
paper2_steadystate.m (2.0 KB)

7

Thank you very much! Problem solved.