[R-sig-ME] Covariance matrices in GLMER

Voeten, C.C.

2018-10-30 13:17:17 UTC

Dear Diana,

1. No, both lmer and glmer by default use unstructured covariance matrices, although they can be massaged into using diagonal, see afex::lmer_alt. Those are the only two options.
2a. Support for other covariance matrices is available in MASS::glmmPQL (which I would not recommend; it inherits all the limitations of nlme::lme and uses PQL) and in glmmTMB::glmmTMB. Depending on what exactly you need to do, glmmTMB is probably what you're looking for.
2b. For glmer and glmmTMB, the ranef function will probably give you what you need.

Best,
Cesko

-----Oorspronkelijk bericht-----
project.org] Namens Kornbrot, Diana
Verzonden: dinsdag 30 oktober 2018 10:16
Onderwerp: [R-sig-ME] Covariance matrices in GLMER
It seems to me that GLMER always uses diagonal matrices.
1. Is this correct?
2. If not, (a) how does one specify different covariance matrices (b) how
does one get output of covariance matrices that GLMER actually used/
Thanks for all help Diana _____________________________________
Professor Diana Kornbrot
Mobile
+44 (0) 7403 18 16 12
Work
University of Hertfordshire
College Lane, Hatfield, Hertfordshire AL10 9AB, UK
+44 (0) 170 728 4626
http://dianakornbrot.wordpress.com/
http://go.herts.ac.uk/Diana_Kornbrot
skype: kornbrotme
Home
19 Elmhurst Avenue
London N2 0LT, UK
+44 (0) 208 444 2081
------------------------------------------------------------
[[alternative HTML version deleted]]
_______________________________________________
https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models