Message-ID: <9C6F0C11872C98439322A20E9DB369F803D9F41B@IS-EX-BEV2.unimelb.edu.au>
Date: 2008-12-11T23:25:51Z
From: Kris Jamsen
Subject: xtable question
In-Reply-To: <40e66e0b0812111453n69f120cayb9385084e8c0d1ac@mail.gmail.com>
This looks promising - I'll give it a go.
Cheers,
Kris
-----Original Message-----
From: dmbates at gmail.com [mailto:dmbates at gmail.com] On Behalf Of Douglas
Bates
Sent: Friday, 12 December 2008 9:54 AM
To: Kris Jamsen
Cc: Martin Henry H. Stevens; r-sig-mixed-models at r-project.org
Subject: Re: [R-sig-ME] xtable question
On Thu, Dec 11, 2008 at 4:25 PM, Kris Jamsen <kmjamsen at unimelb.edu.au>
wrote:
> Thanks Hank for your quick reply. Yes, I tried to coerce my nlme
object
> into a data.frame object but xtable told me I couldn't do that. The
> nlme object is just an nlme model, so I would imagine it would be
> similar in nature to an lm object, but with a few more bits. It looks
> like I might have to try to extend the list of available classes in
> xtable (this might take some time!) or do something more manual.
Would
> be glad to hear more suggestions!
An nlme object is not just an extension of an lm object. There is a
lot more going on in nlme or lme than is the case in lm.
I expect that in creating a table from such an object you want the
table of estimates of the fixed-effects parameters, their approximate
standard errors and so on displayed as a table. To get that table you
apply summary to the fitted model and save the result. There is a
component of that object with a name like ttable which is the table of
fixed-effects estimates, etc. Use
str(summary(lmeFittedModel))
to find the exact name and format.