joergrieger / bvars Goto Github PK

Hi,

the historical decomposition function does not work using the example in the document. the fevd worked, but the output I cannot understand it.

Any help in relation to those I will appreciate it.

Dear Joerg,
Thank you very much for the code "bvars", very clear and user friendly.
Still I would like to clarify one question: when I execute the code (msvar) with the example data UsMonPol and then draw impulse responses, there is a very slight difference in the plots for 2 regimes. It seems like the pictures are pretty the same. I also tried it with my data and got the same result.
What actually influences the difference in impulse responses between the regimes? Could it be a bug in a code?
Thanks in advance!

When intercept == FALSE, should it be modified to y <- c(t(y)) ?

Dear Joerg,

Thank you so much for providing your codes.
I would like to inquire if there is a way to extract individual IRFs from Bayesian VAR estimation?

Kind regards,
Vanessa

Hello, I hope you are well.

I am interested in using the the library to estimate a BVAR using sign restrictions and produce IRF, FEVD and perhaps Historical Decomposition analysis (unconstrained in this case). For VAR estimation and IRF analysis, I was able to do it with the repository documentation and examples provided in https://www.jstatsoft.org/article/view/v100i14. But I can't find a guide for the fevd() and hd() functions and outputs. For example, I am able to generate a fevd object, but I am having problems trying to graph it. And hd function is not recognized in R, could it be that it is on Github and not the imported version by R?

Do you know of any available resource?

Thanks and kind regards,
Franco

I have found my errors. Please delete this issue. Thank you!

First thanks for an excellent package, however I just have one question not necessarily to report a bug but to understand how the cforecast function works in your package. Particularly cfconds, while everything else is clear to me that part still has me confused.

Is there any way you can provide a quick example on it please?

Thanks in advance.

Hi Joerg, thanks for the incredible code you put on here. Especially when it comes to irfs in tvars and other nonlinear models.

However, I had some trouble running diags codes regarding tvars. I couldn't run diag_geweke, trace and stabletest.

Are they really implemented in your package? Am I missing some potentially stupid point?

Thank you very much!

Luckas.