Skip to content
Prev 15993 / 20628 Next

glmmTMB and ar1

Without looking too closely at this, I think you also need to include
(1|Common.Name) in the model; otherwise the assumption is that there is
no within-group variance?

This formula is taken from
http://bbolker.github.io/mixedmodels-misc/notes/corr_braindump.html

glmmTMB_simple_fit <- glmmTMB(y~1 + (1|f) + ar1(tt-1|f),
data=d,family=gaussian)

(you used +0 rather than -1 to suppress the intercept in the ar1() term,
and you have a non-trivial fixed-effect term, but otherwise these are
similar).

  Suggestions for doc improvements/pull requests welcome ...

  Ben Bolker
On 17-11-22 05:26 AM, Jarrod Hadfield wrote: