Log-linearization

[sirene]

Dear all,

I have a confusion about how to make 'the model log-linear. And if we have the log linearized model in our hand, how should we define a vari able ?

Thank you very much in advance,

[jpfeifer]

A variable is a variable and a parameter a parameter, regardless of the model being linearized or not. It seems your confusion derives from defining a model in general. Please consult the user guide and the examples delivered with Dynare.

[sirene]

Thank you very much for your reply.

[sirene]

After looking at the other questions, I would like to ask why we should write a variable as parameter when the model is log-linear? I havent read it in the manual.

Thank you in advance,

[jpfeifer]

I am not saying you should write a variable as a parameter. I am only saying that in log-linearization you sometimes need steady state values or other coefficients as parameters in front of the variables and you must not treat those coefficients/parameters as variables.

[selima]

In fact in the same topic I consider some initial value to linearize the model but I meet this problem :

Starting Dynare (version 4.3.2).
Starting preprocessing of the model file ...
ERROR: NK_GM052C.mod:172.5-10: y is not a parameter

??? Error using ==> dynare at 114
DYNARE: preprocessing failed


what is the problem?

[jpfeifer]

You wrote intival instead of initval