On 27.12.2011 14:43, Silvano wrote:
Hi,
I'm analyzing a longitudinal data set with 387 cows were
observed in 63
days divided into 6 groups, and every 30 days was found
to produce milk.
Does not aim to model the time using regression. Only
compare the groups
differ in terms of milk production. There are many
missing observations.
Because the data are correlated I used the SAS program:
proc mixed data=univar method=reml;
class RACA GRUPO APELIDO Dias;
model Prod = GRUPO / solution DDFM=BW;
repeated Dias / type=arh(1) subject=APELIDO r rcorr;
lsmeans GRUPO / pdiff adjust=tukey;
run ;
See package SASmixed that includes some nice examples how
to move from SAS to R.
Uwe Ligges
But, I want use R. What would be the equivalent in R?
Thank you.
--------------------------------------
Silvano Cesar da Costa
Departamento de Estat?stica
Universidade Estadual de Londrina
Fone: 3371-4346