[S] Retrieving objects from backup

Carmen Maria Cadarso (eicadar@usc.es)
Tue, 17 Feb 1998 19:57:36 +0100


Hi everybody,

I am using the Oswald package, version 3.2, under Splus 4.2.
At this moment, I am working with longitudinal data set with 336 series
(as.balanced) and only a factor covariate. There are no missing values.

1) I used mixed procedure and obtained good results. Nevertheless
a)I couldn't represent the fitted values.
b) I didn't find the chitest and anova test based on mixed models.
It is no possible?

2) Instead of the procedure mixed, I tried the pcmid procedure, but

a) In case of interaction it was no possible to find a solution (i waited
for it several hours)
b) I didn't find chitest or anova test for pcmid models...

What strategy do you recommend? When I used Oswald 3.1 (and the old
version of Splus) I found solutions trough reml.fit procedure...

Best wishes,

Carmen Cadarso