[R-sig-ME] [R] lme nesting/interaction advice

John Maindonald john.maindonald at anu.edu.au
Mon May 12 02:11:40 CEST 2008

This leads to the notion of a fully cross design.  A fully cross
design has the characteristics that:
(a) its lineaments are not clear.
(b) it leads to heated discussion.

John Maindonald             email: john.maindonald at anu.edu.au
phone : +61 2 (6125)3473    fax  : +61 2(6125)5549
Centre for Mathematics & Its Applications, Room 1194,
John Dedman Mathematical Sciences Building (Building 27)
Australian National University, Canberra ACT 0200.

On 12 May 2008, at 7:45 AM, Andrew Robinson wrote:

> On Sun, May 11, 2008 at 07:52:50PM +0100, Federico Calboli wrote:
>> The main point of my question is, having a 3 way anova (or ancova, if
>> you prefer), with *no* nesting, 2 fixed effects and 1 random effect,
>> why is it so boneheaded difficult to specify a bog standard fully
>> crossed model? I'm not talking about some rarified esoteric model
>> here, we're talking about stuff tought in a first year Biology Stats
>> course here[1].
> That may be so, but I've never needed to use one.
> If it's bog-standard and yet boneheaded difficult, then presumably
> someone else would have had this problem before you.  Perhaps a search
> of the archives will help?  If you try, you will find many qualifiers
> to the effect that "lme isn't very well set up for crossed random
> effects".
>> Now, to avoid any chances of being misunderstood in my use of the
>> words 'fully crossed model', what I mean is a simple
>> y ~ effect1 * effect2 * effect3
>> with effect3 being random (all all the jazz that comes from this
>> fact). I fully apprecciate that the only reasonable F-tests would be
>> for effect1, effect2 and effect1:effect2, but there is no way I can
>> use lme to specify such simple thing without getting the *wrong*
>> denDF. I need light on this topic and I'd say it's a general enough
>> question not to need much more handholding than this.
> Perhaps there are some circumstances unique to your situation.
>> I fully apprecciate that R is developed for love, not money,
> ... as is the R-help community ...
>> and if I
>> knew how to write an user friendly frontend for nlme and lme4 (and I
>> knew how to actually get the model I want) I'd be pretty happy to do
>> so and submit it as a library. In any case, I feel my complaint is
>> pefectly valid, because specifying such basic model should ideally
>> not such a chore, and I think the powers that be might actually find
>> some use from user feedback.
> This is not feedback.  It is a compliant.  But, the complaint boils
> down to the fact that you don't know what you're doing, and you show
> no evidence of having searched the R-help archives.  How is that
> helpful?
>> Once I have sorted how to specify such trivial model I'll face the
>> horror of the nesting, in any case I attach a toy dataset I created
>> especially to test how to specify the correct model (silly me).
> Well, these data seem to differ.  Is replica block?  If not, then how
> can we reproduce your results?  And, if I assume that it is, then the
> output df differ from what you sent in your original mail.  So, I find
> this confusing.
> Then, from your original mail,
>> The easiest model ignores the nested random effects and uses just
>> selection, males and replica and the relative interactions. The
>> model
>> lme(y ~ selection * males, random = ~1|replica/selection/males,  
>> mydata)
> forgive me, but I seem to see nesting in the random statement.  That  
> is
> what happens when we separate factors with a '/'; they are nested.  We
> would expect that statement to not provide the correct df for the
> bog-standard fully crossed design.
> Perhaps if you were to comply with the request at the bottom of each
> R-help email, and provide commented, minimal, self-contained,
> reproducible code, that actually ran, ideally with fewer value
> judgements, you might get more attention from the people who are
> smarter than you and me, but have less time than either of us.
> Andrew
> -- 
> Andrew Robinson
> Department of Mathematics and Statistics            Tel:  
> +61-3-8344-6410
> 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/
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models

More information about the R-sig-mixed-models mailing list