Dynare forums

Hi,

I have the usual problem that I have too many eigenvalues greater than 1 than the number of forward looking variables. One eigenvalue is complex, so that might be the problem. I tried using model_diagnostics(M_,options_,oo_) after running dynare to see what the problem is but I get nothing (completely blank screen). All the inputs for the model_diagnostics function exist after I've run dynare, so do you have any idea of what hte problem would be? My mod file is attached, but I initialize variables using a .mat file and I'm not able to upload it directly, so that's in the zip file.

Thanks!

If there are no problems that model_diagnostics is able to detect, there is no output (this will change in the future).

Complex eigenvalues are common and usually not the source of problems (unless your IRFs start oscillating).

As always when the BK-conditions are not satisfied, check the timing and the parametrization.

Thanks.

Hi,

I found a timing issue and sorted it out, and simplified the model a bit. I now get this error:
There are 31 eigenvalue(s) larger than 1 in modulus
for 31 forward-looking variable(s)

The rank conditions ISN'T verified!

When I run model diagnostics, though, I get the following:
model_diagnostics(M_,options_,oo_
|
Error: Expression or statement is incorrect--possibly unbalanced (, {, or [.

I've been through the file, and there are no unbalanced brackets. In general I thought Dynare doesn't even run if there were unbalanced brackets. Is it a problem with the Jacobian but this is the output associated with that in model_diagnostics? I looked at the code here:
http://www.dynare.org/dynare-matlab-m2h ... stics.html

and I don't see what error would trigger exactly the message I received. Would you be able to help with this? (I'm trying to understand the source of the error, if it is actually the Jacobian or if I have a Walras law problem or if it is actually an unbalanced bracket)

Thanks

From what I see, you are missing the closing bracket of

Code: Select all

model_diagnostics(M_,options_,oo_)

Oh, sorry. That was pretty stupid. Thanks anyway, have a good weekend.