R help - Feb 2018 - Specify multiple nested random effects in lme with heteroskedastic variance across group

I want to fit a random effects model with two separate nested random
effects. I can easily do this using the `lmer` package in R. Here's how:

    model<-lmer(y ~ 1 + x + (1 | oid/gid) + (1 | did/gid), data=data)

Here, I'm fitting a random intercept for `oid` nested within `gid` and
`did` nested within `gid`. This works well. However, I want to fit a model
where the variance of the intercept changes with the `gid` for both the
random effects. `nlme` package is capable of doing that. However, it's not
clear how. The best I could do is like so:

    model <- lme(y ~ 1 + x, random=list(gid=~1, oid=~1, did=~1),
weights=varIdent(form=~1|gid), data = data)

but this nests the `did` within `oid` and `gid` nested together. I tried to
use the idea from a similar [question][1], which seems like a close problem
but the answer has not been explained well in that question. I hope someone
can figure this out.


  [1]:
https://stats.stackexchange.com/questions/58669/specifying-multiple-separate-random-effects-in-lme

	[[alternative HTML version deleted]]

You should post this on the r-sig-mixed-models  list. You are more likely
to get a helpful response there.


Cheers,
Bert



Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )

On Tue, Feb 20, 2018 at 8:06 PM, suraj keshri <surajkeshri at gmail.com>
wrote:
> I want to fit a random effects model with two separate nested random
> effects. I can easily do this using the `lmer` package in R. Here's
how:
>
>     model<-lmer(y ~ 1 + x + (1 | oid/gid) + (1 | did/gid), data=data)
>
> Here, I'm fitting a random intercept for `oid` nested within `gid` and
> `did` nested within `gid`. This works well. However, I want to fit a model
> where the variance of the intercept changes with the `gid` for both the
> random effects. `nlme` package is capable of doing that. However, it's
not
> clear how. The best I could do is like so:
>
>     model <- lme(y ~ 1 + x, random=list(gid=~1, oid=~1, did=~1),
> weights=varIdent(form=~1|gid), data = data)
>
> but this nests the `did` within `oid` and `gid` nested together. I tried to
> use the idea from a similar [question][1], which seems like a close problem
> but the answer has not been explained well in that question. I hope someone
> can figure this out.
>
>
>   [1]:
> https://stats.stackexchange.com/questions/58669/
> specifying-multiple-separate-random-effects-in-lme
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/
> posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]