How to run an Auto Regressive Generalized Linear Mixed Model

2 messages · Danielle Rudley, Ben Bolker

Wed, Jun 23, 2021 3:25 AM #

Hello all,

I'm trying to run Auto Regressive Linear Mixed Models in R using the nlme()
package. For my continuous response variables I have used lme() with the
correlation set as 'corAR1()', but one of my response variables has a
poisson distribution (i.e.- count of number of conductive xylem cells). I
was wondering how to run an AR Generalized LMM within the nlme() package,
or if there was another solution to this. Thanks so much for your help.

Cheers,

Danielle Rudley

PhD Candidate
Forest Research Centre (CEF)
School of Agriculture, University of Lisbon
Tapada da Ajuda, 1349-017 Lisboa, Portugal

	[[alternative HTML version deleted]]

Ben Bolker

Wed, Jun 23, 2021 6:27 AM #

You can use the glmmPQL function in the MASS package; the glmmTMB 
package can also handle autoregressive GLMMs. (Also, MCMCglmm and brms 
if you want to go Bayesian.)

On 6/23/21 6:25 AM, Danielle Rudley wrote: