A problem about Bayesian estimation, Thx for reading!

[lssp]

Hello Mr. Pfeifer!

Thank you so much for reading this post!

I've got a problem : After I finished my bayesian estimation in Dynare, how can I get the plot of some certain endogenous variable？

For example, the equilibrium real interest rate is one of the endogenous variables in my model, I use the data between 1962-2008 to do the bayesian estimation, then what can I do to
get the plot of this equilibrium real interest rate between 1962-2008 (just like the picture attached) ?


Would you please help me out ? Thanks again!

[jpfeifer]

You need to specify the

Code: Select all

smoother

option.

[lssp]

You need to specify the

Code: Select all

smoother

option.

Professor Pfeifer,

Thank you so much for your reply, that really helps me a lot ! And after I add " smoother " in the code, I've got some new questions, would you please help me out ?

(1) Affter I add the "smoother", I sucessfully get the plots of all the variables, what if I only want to get the plot of one certain variable (e.g. the interest rate) ? What should I suppose to do ?

(2) Then it's about the Y axis. The model is in linear form, so the value of the variable (e.g. the interest rate) displayed in Y axis is the value after log-linearized, how can I change it into the real value and make it displayed in the Y axis ?

(3) If I just want to get the mean estimate of the smoothed variables ( the black line ) in the plot and do not need the deciles of the smoother distribution (the green lines), what should I do to eliminate the green lines ?

Thank you again for reading this post and it's very kind of you for solving these problems !

[jpfeifer]

1) Only provide that variable after the estimation command
2) Define a new variable that takes the level. If your model is loglinearized, this might be complicated.
3) You need to generate this graph yourself using the results saved in oo_.SmoothedVariables