[R-sig-ME] Random-effects model with known error variance

[Doran, Harold]

Yes, lmer can handle unbalanced data. No, you cannot constrain one of the variances to have a fixed value using lmer. Solutions for linear mixed models are well-known, and so if your grasp of R programming and matrix algebra is decent, then it is very easy to do this.  

> -----Original Message-----
> From: r-sig-mixed-models-bounces at r-project.org [mailto:r-sig-mixed-models-
> bounces at r-project.org] On Behalf Of Asaf Weinstein
> Sent: Sunday, November 11, 2012 11:08 PM
> To: r-sig-mixed-models at r-project.org
> Subject: [R-sig-ME] Random-effects model with known error variance
> 
> Hi there,
> 
> I am trying to fit a random effects two-way linear model with unbalanced data,
> and have two questions:
> 
>    1. Can someone verify that lmer() can indeed cope with an unbalanced
>    table?
>    2. This might be more difficult: I am looking to fit the same model with
>    a KNOWN value for the error variance (sigma sq). I wonder if this is
>    possible to do with current package, and if not, if it's a good idea to
>    look into the implementation and try to do it myself. I'm curious to know
>    how difficult that would be to a not-too-sophisticated R user.
> 
> Thanks so much,
> 
> Asaf
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-mixed-models at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models