question about the posterior intervals of the local variable
Posted:
Thu May 04, 2017 6:39 amby zhanshuo
Hi, dear professor,I encounter a problem of calculating the posterior intervals of parameters, which denote as model-local variables by"#" in the model section.
So is there a easy way to work it out?
Thanks.
Re: question about the posterior intervals of the local vari
Posted:
Thu May 04, 2017 11:32 amby jpfeifer
That does not work with model local variables. What you can do instead is to define the model local variable as a parameter and handle the parameter dependence via a steady_state_model-block.
Re: question about the posterior intervals of the local vari
Posted:
Thu May 04, 2017 1:04 pmby zhanshuo
Thanks for answering me!
But I still doubt since steady-state-model is a file that handles endogenous variables, but my model-local variable is a parameter, like
"# r_ss=1/beta", where beta is a estimated parameter and thus in this way r_ss is also estimated.
So can such a parameter be defined in the steady-state-model?
Thanks!
Re: question about the posterior intervals of the local vari
Posted:
Fri May 05, 2017 6:55 amby jpfeifer
Yes. For every iteration, the steady_state_model-block will be evaluated. Any parameter you set within the steady_state_model-block or the steadystate.m-file will be passed back. That allows handling parameter dependence as well. An example of this is at
https://github.com/JohannesPfeifer/DSGE_mod/blob/master/Gali_2015/Gali_2015_chapter_6_5.mod