On Thu, 22 Oct 2009, [iso-8859-1] Javier Garc�a wrote: > Nowadays the fractional methods are more and more used, and as > Gretl has implemented the ADF and ADF-GLS tests, I think it > would be very interesting to complement them with the fractional > version of the Dickey Fuller test (Dolado, Gonzalo and Mayoral > -Econometrica, 2002 & Studies in Nonlinear Studies and > Econometrics 2008-; and Lobato and Velasco -Econometrica, > 2007-).
> Have anybody programed it? Not that I know of. > I have some routines, but there are in R and they are quite > "rudimentary". Hovewer, if you are interested in this test I > could pass you my routines so as you can review and implement if > you want. You're welcome to send the R code and we can see about implementing it in gretl (though I'm not promising it'll happen right away). Another (perhaps better) option would be to write and submit a gretl function package that does the job. See Chapter 10 of the Gretl User's Guide, section 10.6 in particular. Allin Cottrell