Dear Ladies and Gentlemen,
currently I'm struggling with the analysis of several datasets. They have one thing in common - simple nested or crossed random effects do not correctly represent the dependency matrix of my random effects. It was suggested that pdMat dependency structures in nlme may be very suitable to describe the dependency structures of my random effects. However, in most standard statistical literature, this topic is hardly mentioned. Even when it is mentioned somewhere, it covers barely half a page and for the most part only lists what the function is capable of. What I'm looking for is a comprehensive but non-mathematical approach that teaches one to choose the correct dependency structure for several examples. Do you maybe know of a suitable reference or would anyone be willing to teach me how to do that in a personal conversation?
Thank you for your time!
Denis
The definitive reference for nlme is
Pinheiro, J. C. & Bates, D. M. Mixed-effect models in S and S-Plus. Springer, 2000, 528 p.
See https://www.researchgate.net/publication/208033930_Mixed-Effect_Models_in_S_and_S-plus
and contact me privately if you fail to find the book there.
The book contains many examples of ad-hoc pdMat matrices. You should also check R-Sig-Mixed archives: see e.g http://marc.info/?l=r-sig-mixed-models&w=2&r=1&s=pdmat+lme&q=b
Book Mixed-Effect Models in S and S-plus
- Badges
- Science topic