These days I use SAS mixed model to compute the variances.( For example, proc mixed data=phe convf=1e-8 maxiter=50; )Through google, I know some R packages, like lme4 and nlme. But I can not get the results like SAS. Now I want to know the corresponding R package having the similar function with SAS mixed model. Thank you very much!
Hi Sofia, thank you for your answer. And you provide me a good material. I will study lme4 again. Maybe there are some functions in lme4 that I do not know. My model is : y = XB + Z1(sigma1)2 + Z2(sigma2)2 + I(sigmae)2, Z1 and Z2 is a n*n matrix respectively, and I is a n*n unit matrix, y is a vector, and we have known y, XB, Z1,Z2. Then we want to compute (sigma1)2, (sigma2)2 and (sigmae)2. This is easy for SAS, but it is difficult for R. If you have some advice, please tell me. Thank you again.
Hi Sofia,
I will try to find some relations about mixed model between SAS and R through reading the file. Thank you very much!
You may wish to know about
R2MLwiN is an R command interface to the MLwiN multilevel modelling software package, allowing users to fit multilevel models using MLwiN from within the R environment.
http://www.bristol.ac.uk/cmm/software/r2mlwin/
MLwiN is a fast specialist mixed modeling software that can fit a large range of models in likelihood and MCMC estimation. You can fit models with design matrices but you would have to set them up beforehand.
To be clear you need R and MLwiN on your machine to use this facility.
For a quick guide - see this slide show
http://www.bristol.ac.uk/cmm/media/r2mlwin/Parker_et_al_2015_Utrecht.pdf
Dear prof. Kelvyn,
R2MLwiN is a fresh software for me. And from your description about MLwiN, I think it's the software that I want. Thank you!
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R Statistical Package