VAR impulse response output

[Willem]

Hi,

I try to extract the stdev of accumulated impulse responses.
the program below produces and freezes the output Table, from which I try to extract the stdev.
However, since the stdev is between brackets, I can not take it is a numerical value.
and therefore my program (which stores values from the Table) does not work
the code creates, freezes the Table and reads cell 18,2: the impulse response, or instance: "20450" and cell 19,2: this is the stdev, for instance "(897)"
the problem is that the program can not extract the latter, because of the brackets.
How to solve this ?

var var_dum.ls 1 2 dbal dfe1 @ c
freeze(tabel) var_dum.impulse(8,t,a,se=a,imp=chol) dbal @ dfe1
coef(!t)=tabel(18,2)
conf_low (!t)=tabel(19,2)

[EViews Gareth]

The .impulse command has options to save the impulses (and their SEs) into a matrix object for you.
http://www.eviews.com/help/helpintro.ht ... 23ww219724

[g.ciminelli]

Hi, I estimated a PVAR with 5 endogenous variables, 2 lags, 16 cross sections, about 30 years of observations per cross section, time fixed effects, country fixed effects and countr-specific linear time trends.

I now want to construct my own IRFs (in order to normalize the size of the shocks, change graph layout etc.)

When I try to save responses ordered by shocks in the matrix named prova using the function matbys, I get the following error: "matrix size mismatch". Do you have any idea why this is the case.

Thanks in advance
Gabriele

PS, this is the code line:

var_1_1_2_1.impulse(11,se=mc,rep=1000,matbys=prova) x_1_1_2_1 lgdp_pc lgini_net lab_une llab_par @ x_1_1_2_1

[EViews Gareth]

EViews doesn't estimate PVARs :)

Could you provide your workfile?

[g.ciminelli]

Well, there is not a built in function if that's what you mean.

But you can still estimate a VAR in a panel workfile.

I seem to have solved the issue anyway, thanks for your interest!