[R] fixed effects

[Spencer Graves]

	  Have you tried the following:

	  lme(y~1, random=~1|X, data=DF)

where DF = a data.frame with columns y and X.

	  The authoritative reference on library(nlme) is Pinheiro and Bates 
(2000) Mixed-Effects Models in S and S-Plus (Springer).  I've learned a 
lot from Bates and from this book in particular.

	  hope this helps,
	  spencer graves

ivo welch wrote:

> dear R wizards:
> 
> X is factor with 20,000*20=800,000 observations of 20,000 factors. 
> I.e., each factor has 20 observations.  y is 800,000 normally
> distributed data points.  I want to see how much R^2 the X factors can
> provide.  Easy, right?
> 
> 
>>lm ( y ~ X)
> 
> and
> 
>> aov( y ~ X)
> 
> Error: cannot allocate vector of size 3125000 Kb
> 
> is this computationally infeasible?  (I am not an expert, but
> off-hand, I thought this can work as long as the X's are just factors
> = fitted means.)
> 
> 
>>help.search("fixed.effects");
> 
> 
> fixed.effects(nlme)                 Extract Fixed Effects
> fixed.effects.lmList(nlme)          Extract lmList Fixed Effects
> lme(nlme)                           Linear Mixed-Effects Models
> lmeStruct(nlme)                     Linear Mixed-Effects Structure
> nlme(nlme)                          Nonlinear Mixed-Effects Models
> nlmeStruct(nlme)                    Nonlinear Mixed-Effects Structure
> 
> Type 'help(FOO, package = PKG)' to inspect entry 'FOO(PKG) TITLE'.
> 
> ok---I want to read the fixed.effects help.  could the help system
> tell me how to inspect these entries?
> 
> 
>>help("fixed.effects")
> 
> wrong
> 
>>help.search("fixed.effects")
> 
> two entries, as above.  nothing new.
> 
>>?fixed.effects
> 
> wrong
> 
> eventually, it dawned on me that nlme in parens was not a function
> argument, but the name of the package within which fixed.effects
> lives.  Suggestion:  maybe a different notation to name packages would
> be more intuitive in the help system.  yes, I know it now, but other
> novices may not.  even a colon instead of a () may be more intuitive
> in this context.
> 
> 
>>library(nlme); ?lme
> 
> and then
> 
>>lme(y ~ X)
> 
> Error in getGroups.data.frame(dataMix, groups) :
>         Invalid formula for groups
> 
> 
> now I have to beg for some help.  ok, blatant free-riding.  the lme
> docs tells me it does the Laird and Ware model, but I do not know this
> model.  the only two examples given at the end of the lme help file
> seem to be similar to what I just specified.  so, how do I execute a
> simple fixed-effects model?  (later, I may want to add a variable Z
> that is a continuous random variable.)  could someone please give me
> one quick example?  help is, as always, highly appreciated.
> 
> sincerely,
> 
> /ivo welch
> 
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html