[RsR] Package robustlmm

[Manuel Koller]

Dear Mina

Sorry for the late reply.

These random errors are usually due to the system running out of memory.

Your data.frame A contains over 58'000 observations. The rlmer method
requires more complex equations than lmer, which allow for less memory
efficiency. Maybe one could do more with sparse matrices, as your data
seems to be factors only. Everything depends on how filled up the
intermediate matrices actually are. So this would be a long shot.

I would say your best bet is trying to get access to a machine with a lot
of memory. You will need at least 64GB of ram. Just the matrix that is
evaluated at the point of failure requires a bit more than 25GB.

Best,
Manuel




On Wed, 4 Aug 2021, 20:14 Matias Salibian-Barrera, <matias using stat.ubc.ca>
wrote:

> Hello Mina,
>
> I tried to reproduce your error but couldn't so far because the line
>
> fit.robustlmer<-rlmer(dbp ~ age+sex+bmi+time+(1|id),data=A)
>
> is taking too long on my laptop.
>
> I suggest first using traceback() to try to narrow down where the error
> occurs, and then contacting the package maintainer with a smaller
> reproducible example if possible.
>
> Sincerely,
>
> Matias
>
> ______________________________________________________________
> Matias Salibian-Barrera - Professor
> Department of Statistics - The University of British Columbia
> matias using stat.ubc.ca | +1 (604) 822-3410
>
>
> On 2021-08-04 2:18 a.m., mina jahan wrote:
> > [*CAUTION:* Non-UBC Email]
> >
> > Hi,
> >
> > I want to use the rlmer function of package robustlmm on the real
> > longitudinal data, but I was exposed with this error:
> > Error in.local (x,y,...) : negative length vectors are not allowed.
> >
> > I sent R code (txt file) and data (csv file).
> > I apologize for any inconvenience that my request may cause. I look
> > forward to receiving your reply.
> >
> > Thanks a lot in advance for your response.
> >
> >
> > cheers
> > Mina
> >
> > _______________________________________________
> > R-SIG-Robust using r-project.org mailing list
> > https://stat.ethz.ch/mailman/listinfo/r-sig-robust
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-SIG-Robust using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-robust
>

	[[alternative HTML version deleted]]