[R-sig-ME] non-parametric mixed model

[David Duffy]

On Wed, 1 Feb 2012, Lene Jung wrote:

> variables. For all variables, I have repeated measurements from individuals
> (multiple seasons) and 2 different study sites, so I am using a mixed-model
> approach with individual as subject (random effect) to test for differences
> between sites and seasons.
>
> My problem is that I have 2 variables that are cosines which means that
> they are in the interval -1 to 1. One of the variables have many ties (many
> 1s) whereas the other doesn't.

This is circular data?  There apparently are mixed models eg for 
longitudinal circular data, and there are nonparametric tests of 
homogeneity.

eg http://www.omicsonline.org/2155-6180/2155-6180-1-107.php,
R CircStats package.

Or am I completely misunderstanding?

-- 
| David Duffy (MBBS PhD)                                         ,-_|\
| email: davidD at qimr.edu.au  ph: INT+61+7+3362-0217 fax: -0101  /     *
| Epidemiology Unit, Queensland Institute of Medical Research   \_,-._/
| 300 Herston Rd, Brisbane, Queensland 4029, Australia  GPG 4D0B994A v