Page 1 of 1

VAR Data Members

Posted: Tue Nov 05, 2013 10:36 am
I need to save the Schwarz criterion and the log likelihood of a specific equation of a VAR.

This for example doe not work:
var example_var.ls 1 2 x1 x2 x3
scalar loglikelihood = example_var.@eqlogl(1)
scalar SC = example_var.@schwarz(1)

Re: VAR Data Members

Posted: Tue Nov 05, 2013 10:49 am
VAR data members are located at the start of the VAR section of the Object Reference.

Each object's data members are located at the start of its section in the Object Reference.

Re: VAR Data Members

Posted: Tue Nov 05, 2013 11:10 am
Yes, I looked at the Object reference. As said this does not worK:

var example_var.ls 1 2 x1 x2 x3
scalar loglikelihood_x1 = example_var.@eqlogl(1)
scalar SC_x1 = example_var.@sc(1)

The numbers I get are not those reproduced by EVIEWS in its Vector Autoregression Estimates.

Re: VAR Data Members

Posted: Tue Nov 05, 2013 11:25 am
Ah, sorry, I missed the part where you said you wanted the equation specific one. Only the system wide ICs are available from the VAR, not the equation specific ones.

You can, of course calculate it though:

Code: Select all

`create u 100series x1=nrndseries x2=nrndseries x3=nrndvar example_var.ls 1 2 x1 x2 x3scalar loglikelihood_x1 = example_var.@eqlogl(1)!t = example_var.@eqregobs(1)!l = example_var.@eqlogl(1)!n = example_var.@eqncoef(1)scalar SC_x1 = -2*!l/!t + !n*@log(!t)/!tshow example_varshow sc_x1`

Re: VAR Data Members

Posted: Tue Nov 05, 2013 11:35 am
These are the number of your VAR
x1 x2 x3
Log likelihood -144.4427 -132.9798 -138.4324
Schwarz SC 3.275307 3.041371 3.152648

and these are the values of your scalar sc_x1 = 8.78354 and loglikelihood_x1 = -414.34622

They are very different.

Re: VAR Data Members

Posted: Tue Nov 05, 2013 11:41 am
In mine they match. What's the version and build date of your copy of EViews (Help->About EViews)?

Re: VAR Data Members

Posted: Wed Nov 06, 2013 1:43 am
Version 7.2
Enterprise edition Sep 23 2011 buid

Re: VAR Data Members

Posted: Wed Nov 06, 2013 7:58 am