Re: [Gretl-users] acf and pacf output
On Fri, 2 Apr 2004, Allin Cottrell wrote:
2. Gretl's PACF numbers used the estimator that is set out in
several
standard sources (Greene, Hamilton, Davidson and MacKinnon): namely,
the coefficient for lag k is the last coefficient in an OLS regression
of y_t on [1, y_{t-1}, y_{t-2}, ..., y_{t-k}]. However, it appears
that the procedure used by R and Eviews is the Durbin-Levinson
algorithm, which computes the sample PACF recursively from the sample
ACF values. I have now switched to Durbin-Levinson. (From my reading
on this topic, I'm not quite clear on why the results differ: any
insights into that?)
One first guess would be that the sample is different. Basically, PACF(k)
is a function of ACF(1)..ACF(k); these could be computed, respectively, on
a sample of size (T-1), (T-2), ... (T-k), as I suspect R does, or (as
would happen in a regression), by simply discarding k observations for
each one. Asymptotically, it wouldn't matter.
Riccardo `Jack' Lucchetti
Dipartimento di Economia
Università di Ancona
jack(a)dea.unian.it
http://www.econ.unian.it/lucchetti