Hi Guiseppe,
On Wed, Nov 29, 2006 at 01:26:37AM +0000, giuseppe scasserra
wrote:> hi, my name is giuseppe and i am an engineer, i am using R for regressions
> analysis using mixed random effects model, funztion "nlme".
> quastions are:
> 1) when i ask for the summary after a regression using nlme function the
> program gives me the values of the coefficients and also the standard
> deviation of the random effect but if i ask for the function
> sd(random.effects(model)) the value is different, why?
My understanding is that (and I welcome comments), the summary
function reports the maximum likelihood (or residual maxiumum
likelihood) estimate of the standard deviation of the random effect.
The value computed by sd(random.effects(model)) is the method of
moments estimate of the standard deviation of the random effect.
These can easily be different. They are not constrained to be the
same.
> 2) which is the real meanings of the function "residuals" in the
nlme
> function ? (analitically)
I am not sure what you mena by the analytical meaning, so I will not
comment.
> 3) have you got any suggestion about books or papers about nlme function in
> R ?
Absolutely - Pinheiro and Bates 2002 is essential reading. Following
is a link to the book information on Amazon
http://www.amazon.com/Mixed-Effects-Models-S-S-Plus/dp/0387989579/sr=1-1/qid=1164764746/ref=pd_bbs_sr_1/103-6675834-9672605?ie=UTF8&s=books
Cheers
Andrew
--
Andrew Robinson
Department of Mathematics and Statistics Tel: +61-3-8344-9763
University of Melbourne, VIC 3010 Australia Fax: +61-3-8344-4599
http://www.ms.unimelb.edu.au/~andrewpr
http://blogs.mbs.edu/fishing-in-the-bay/