I've got some code block from Riccardo which does a perfect job, Sven.

Indeed, the alternative would be the step-by-step pretesting using the "omit --auto" function. However, as far as I understood their are some issues using step-wise pretesting...

Artur

2015-03-25 14:25 GMT+01:00 Sven Schreiber <svetosch@gmx.net>:
Am 25.03.2015 um 11:05 schrieb Artur T.:
Dear all,

I want to find the optimal lag length of an ARDL model allowing for
variable-specific lag lengths.

For instance, I've got K variables and want to select the optimal lag
structure by searching across the (pmax-1)^K ARDL models, spanned by
p=0,1,..., pmax and p_k=0,1,..., pmax, k=1,.., K using some criteria
(e.g. AIC).

I am looking for a clever trick to estimate all of these possible models
but only have a complicated "loop-in-loop-in-loop" structure in mind.
Does anybody have a clever idea how to realize this task efficiently?

Yes and no -- No if you want a general solution, because then in my mind there's basically no way around an exhaustive nested loop. Yes if you accept some compromise and possible path-dependence, because then it seems you could just use "omit --auto", no?



Btw, it seems that the latest Eviews 9 version has this feature included:
http://www.eviews.com/EViews9/ev9ecest_n.html#ardl


I didn't even know that Eviews 9 has come out.

cheers,
sven

_______________________________________________
Gretl-users mailing list
Gretl-users@lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users