Declare priors as a function

Estimation ML/Bayesian estimation
1

Is it possible to declare a parameter as a function? For example?

Instead of

Phi, gamma_pdf, 0, 0.1;

Is it possible to declare it as ;

100* (Phi - 1), gamma_pdf, 0.95, 0.1 ?

2

No, in that case, it is better to define

Phi_trans, gamma_pdf, 0.95, 0.1:

and have a model-local variable in the model-block

#Phi=Phi_trans/100+ 1

that tracks the transformation for you.

3

Thank you sir

4

Hello professor,

When I define a model local variable as per above instruction, the steady state file cannot recognize this variable causing Dynare to prompt an error. How do I fix this problem?

5

If you are having a steady state file, you typically handle parameter dependence in that file not via model-local variables.