s-news
[Top] [All Lists]

Can "lme" build a linear mixed effects model without random intercept?

To: s-news@lists.biostat.wustl.edu
Subject: Can "lme" build a linear mixed effects model without random intercept?
From: "Ramen Bu" <iscenario@hotmail.com>
Date: Mon, 18 Feb 2002 23:21:32 -0500


Dear all,

S-plus 6.0 in Win 2000.

I've built a model with one intercept random effect and one time random effect, the computer code as below.

lme1 _ lme ( fixed = Y ~ A + B ,
            random = ~ time | seq,
            na.action = na.omit)

If I want to build a model the same as above, except do not put in the random time effect, I can write the code

lme1 _ lme ( fixed = Y ~ A + B ,
            random = ~ 1 | seq,
            na.action = na.omit)

The question is : Can I build a model only with the time random effect, that is, do not put in the random intercept effect?

Thanks!

Ramen



_________________________________________________________________
Chat with friends online, try MSN Messenger: http://messenger.msn.com


<Prev in Thread] Current Thread [Next in Thread>
  • Can "lme" build a linear mixed effects model without random intercept?, Ramen Bu <=