I answer about my own question...
On Monday 21 May 2007 12:40, Manuele Pesenti wrote:> Dear R User,
> I'm a new member of this list, I'm using R language from a short
time and I
> would like to ask you something I cannot solve...
>
> I'm using the fwdlm function from the forward library, I'm
specially
> interested by the results of Cook's distances: I'd want to know
which
> inputs are the cause of peacks in Cook's distances plot; because what I
> find in abscisas axes are the steps in which some (up to 5) input data are
> introduced to estimate a new LMS regression.
>
> can someone help me?
in value "Unit" of an fwdlm oblect you find the progressive number of
inputs
as they appear in the data set passed to fwdlm function, in this way is
possible to manually assoiate Cook's distances values to inputs value...
someone has an idea how to do it automatically?
thank you
best regards
Manuele PEsenti
>
> thank you very much in advance
> best regards
> Manuele Pesenti
--
Manuele Pesenti
manuele a inventati.org
amicogodzilla a jabber.linux.it
http://mpesenti.polito.it