[R-sig-ME] separate ar1 structures for multiple outcomes in a single model

[Timothy MacKenzie]

Dear Colleagues,

I was wondering if there is a way to fit "ar1" random structures
separately for multiple DVs in a single model?

I made the following two attempts using glmmTMB without success:

1- glmmTMB(y ~ time*DVs + ar1(0 + interaction(time,DVs) | id),
             data = dat)

2- glmmTMB(y ~ time*DVs + diag(0 + DVs | id) +
              ar1(0 + interaction(time,DVs) | id),  data = dat)

where "time" and "DVs" are factors with 4 and 5 levels, respectively.

Thanks,
Tim M