
Hi Jaromir,
I have a panel with T=60 and N=5 with 3 to 5 variables, and I would need to estimate a panel VAR on it.
I was looking at your code, which looks very similar to what I'm doing, but I think I would need to make some modifications to it and not quite sure how 
1) I will need to add dummy variables for different time points in the sample. is this possible in the framework?
2) you say that the model is fixed effects. I thought fixed effects are homogenous so would produce the same IRFs for the N groups? but your results show these as being different across groups  why is this?
thanks a lot in advance!

