[R-sig-ME] corelated errors
Andrew Robinson
A.Robinson at ms.unimelb.edu.au
Mon Nov 26 19:11:30 CET 2007
Hi Iasonas,
it depends on the correlation structure. If you can produce the
structure using crossed or nested random effects, then, yes. If you
are referring to the kinds of correlations that are produced by the
correlation argument in lme(), then, no, not yet.
Cheers
Andrew
On Mon, Nov 26, 2007 at 09:50:22AM -0800, Iasonas Lamprianou wrote:
> Dear friends, may we use lmer to estimate models where residuals can be correlated (no conditional independence
> assumption)
>
> thanks
>
>
> Dr. Iasonas Lamprianou
> Department of Education
> The University of Manchester
> Oxford Road, Manchester M13 9PL, UK
> Tel. 0044 161 275 3485
> iasonas.lamprianou at manchester.ac.uk
>
>
> ----- Original Message ----
> From: "r-sig-mixed-models-request at r-project.org" <r-sig-mixed-models-request at r-project.org>
> To: r-sig-mixed-models at r-project.org
> Sent: Saturday, 10 November, 2007 1:00:01 PM
> Subject: R-sig-mixed-models Digest, Vol 11, Issue 7
>
> Send R-sig-mixed-models mailing list submissions to
> r-sig-mixed-models at r-project.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
> or, via email, send a message with subject or body 'help' to
> r-sig-mixed-models-request at r-project.org
>
> You can reach the person managing the list at
> r-sig-mixed-models-owner at r-project.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of R-sig-mixed-models digest..."
>
>
> Today's Topics:
>
> 1. segmented regression mixed model? (Irene Mantzouni)
> 2. Re: Nested Mixed Models in lme4 (Marco Chiarandini)
> 3. lme4 is now on R-forge (Douglas Bates)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Fri, 9 Nov 2007 16:59:27 +0100
> From: "Irene Mantzouni" <ima at difres.dk>
> Subject: [R-sig-ME] segmented regression mixed model?
> To: <r-sig-mixed-models at r-project.org>
> Message-ID:
> <68E7981938EAF54F987AD3848A0A6416E5837E at ka-mail01.dfu.local>
> Content-Type: text/plain; charset="ISO-8859-7"
>
> Hi all!
>
> Is it possible to use a segmented regression model as the functional form of a linear (or maybe non-linear?) mixed model?
>
> Cheers,
> Irene
>
>
>
> ------------------------------
>
> Message: 2
> Date: Fri, 09 Nov 2007 17:41:34 +0100
> From: Marco Chiarandini <marco at imada.sdu.dk>
> Subject: Re: [R-sig-ME] Nested Mixed Models in lme4
> To: Douglas Bates <bates at stat.wisc.edu>
> Cc: r-sig-mixed-models at r-project.org
> Message-ID: <47348DBE.1060407 at imada.sdu.dk>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Dear Prof. Bates,
>
>
> >> I am trying to use the function lmer from lme4 to
> >> analyse the following nested factorial design.
> >
> >> I have three treatment factors (neighborhood,
> >> initial, k);
> >> I have three group factors crossing (size, dens,
> >> inst).
> >
> > Did you mean to write (size, dens, type) there?
> >
> > Also, by "factor" do you mean that you regard all of these variables
> > as categorical? If so, you should check the form of the size variable
> > in the data frame. It is being stored as a numeric variable, not as a
> > factor. If you want to interpret this variable as a categorical
> > factor you should convert it to a factor or, as seems likely in this
> > case, an ordered factor. (See ?factor and ?ordered)
>
>
> yes, thank you a lot! All your corrections are
> appropriate! inst should have been type and all
> variables should have been categorical. My mistake.
> Also: as you correctly pointed out, the data are
> from a computer experiment and perfectly balanced,
> and by group factors I meant blocking factors.
>
> Your very clear explanation solved my concerns
> about the nesting! Thanks!
>
> I've also redone the comparison with SAS and now
> results correspond.
> The reason was mainly that I needed a quite
> different formula:
>
> lmer(err~initial*neighborhood + initial*k +
> initial*type + initial*size + initial*dens +
> neighborhood*k + neighborhood*type +
> neighborhood*size + neighborhood*dens + k*type +
> k*size + k*dens + type*size + type*dens +
> size*dens + initial*neighborhood*k +
> (1|inst),data=Case3)
>
> True also that we were using lsmeans in SAS that
> you discourage.
>
> To me it would remain only to understand how I
> could obtain the results in a cell means format
> like those in SAS. But this seems to be a problem
> also in lm and hence I must probably study better
> how things work to find the way. Trying something
> of the kind:
>
> fmm1 <-
> lmer(err~-1+ordered(size)+dens+type+(k+initial+neighborhood)^3+(1|inst),data=Case3)
>
> does not seem to help much.
>
> I left all the analysis I did, code + results,
> (SAS and R) at:
>
> http://www.imada.sdu.dk/~marco/Mixed/
>
>
> Thank you a lot very much for the help!
>
> Best regards,
>
> Marco
>
>
>
> --
> Marco Chiarandini
> http://www.imada.sdu.dk/~marco
> Department of Mathematics Email:
> marco at imada.sdu.dk
> and Computer Science, Phone: +45 6550 4031
> University of Southern Denmark Fax: +45
> 6593 2691
>
>
>
> ------------------------------
>
> Message: 3
> Date: Fri, 9 Nov 2007 15:02:52 -0600
> From: "Douglas Bates" <bates at stat.wisc.edu>
> Subject: [R-sig-ME] lme4 is now on R-forge
> To: R-SIG-Mixed-Models at r-project.org
> Message-ID:
> <40e66e0b0711091302h7ccb532bx94c7312526a774b6 at mail.gmail.com>
> Content-Type: text/plain; charset=ISO-8859-1
>
> Users of the current version of the lme4 package have reported several
> problems and, for some time, I have been unresponsive about such reports
> or I have made reference to the development version of the package. Let
> me emphasize that I am grateful for the reports and, indeed, have fixed
> several of these problems in the development version of the package.
> However, I have held off releasing the development version because of
> one small problem - it doesn't fit generalized linear mixed models correctly.
>
> I have had to go back and reformulate the model from scratch so that I
> can understand it and design the code. As anyone who has developed
> and maintained a large project can attest, the only way to build
> trustworthy code (and to maintain your sanity) is to modularize the
> code. It goes without saying that before you can decide how to
> modularize the code you must be able to decompose the steps in the
> computation. The development version is designed to handle linear
> mixed models, generalized linear mixed models, nonlinear mixed models
> and generalized nonlinear mixed models with nested or crossed or
> partially crossed random factors. It has taken me a long time to
> decide how all those pieces fit together. Only in the last couple of
> weeks have I have managed to convince myself that I know how it all
> fits together. The task of convincing others remains, and is
> decidedly non-trivial, but I feel that I can decompose the
> computational steps now.
>
> It will take a while to move from the equations in my lab notebook to
> released code and, during that process, I will probably need to
> reformulate the slots in the S4 classes. My method of getting to the
> final design of the data structures and algorithms is to keep doing it
> wrong 'til I do it right.
>
> So that others have easy access to the development version of the
> package I have moved the repository for the development version of
> the package to http://R-forge.R-project.org/packages/lme4
>
> Martin and I had planned to do this move in a way that would preserve
> the history of the changes from the current repository
> but that is not easy to do because of the way that the Matrix and lme4
> packages were merged then un-merged. Thus I have made a clean break
> and installed the development version (the one known as gappy-lmer) on
> R-forge. You can access it at the URL given above or as
> http://lme4.r-forge.r-project.org/, at the expense of one additional click.
>
> Starting tomorrow you should also be able to install the development
> version of the package with
>
> install.packages("lme4", repos = "http://r-forge.r-project.org")
>
> Please be aware that the class representations can change so when
> using the development version you should not count on being able to re-use
> a fitted model after installing a new version. You should retain the original
> data so you can refit the model if necessary.
>
>
>
> ------------------------------
>
> _______________________________________________
> R-sig-mixed-models mailing list
> R-sig-mixed-models at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>
>
> End of R-sig-mixed-models Digest, Vol 11, Issue 7
> *************************************************
>
>
> ___________________________________________________________
> Want ideas for reducing your carbon footprint? Visit Yahoo! For Good http://uk.promotions.yahoo.com/forgood/environment.html
>
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
--
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/
More information about the R-sig-mixed-models
mailing list