[R-sig-ME] lme(): Error message in the confidence interval

[Christine Adrion]

Hello,

you get this error message when your LME-Model is too complex, i.e. the 
result of estimating a complicated model with very little data.
Generally, if the variance covariance matrix of the random effects is 
not positive definite this could be a sign of a model that is 
inappropriate and not a good match for the available data.
Using the nlme-package, you do not get an error message like this when 
using intervals(my.lme.fit), in contrast to summary(my.lme.fit) where 
this problem remains undetected....

Best,
Christine

r-sig-mixed-models-request at r-project.org schrieb:
> Send R-sig-mixed-models mailing list submissions to
> 	r-sig-mixed-models at r-project.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
> 	https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
> or, via email, send a message with subject or body 'help' to
> 	r-sig-mixed-models-request at r-project.org
>
> You can reach the person managing the list at
> 	r-sig-mixed-models-owner at r-project.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of R-sig-mixed-models digest..."
>
>
> Today's Topics:
>
>    1. Error message in the confidence interval for the
>       autocorrelation (FMH)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Tue, 27 Oct 2009 03:19:11 -0700 (PDT)
> From: FMH <kagba2006 at yahoo.com>
> Subject: [R-sig-ME] Error message in the confidence interval for the
> 	autocorrelation
> To: r-sig-mixed-models at r-project.org
> Message-ID: <393692.97146.qm at web38303.mail.mud.yahoo.com>
> Content-Type: text/plain; charset=iso-8859-1
>
> Dear All,
>
> I have fitted?two different mixed models, with the AR(1) and AR(2) correlation structure, respectively, and tried to find the confidence interval of the intervals of the aurocorrelation estimation i.e. ?phi1 in the first model and phi1 and phi2 in the second model, via 'intervals' command. The first model worked fine as it gives the required confidence interval for phi1, but there is an error message for the second model, as shown below.
>
> ###################################################################################
>   
>> (lme2 <- lme(Temp ~ t1 + t2, data, random = ~ t1 + t2 | group, correlation = corARMA(p = 2, form = ~ t1 | group)))
>> intervals(lme2)
>>     
> ###################################################################################
> ?
> Error in intervals.lme(lme2) : 
> Cannot get confidence intervals on var-cov components: Non-positive definite approximate variance-covariance
> Calls: intervals -> intervals.lme
> Execution halted
> ?
> ###################################################################################
> ?
> Could someone please tell me the meaning of this error and i'd be grateful?for any?advice or suggestion?on the best way to tackle on this problem?
> ?
> Cheers
> Fir
>
>
>     
>
>
> ------------------------------
>
> _______________________________________________
> R-sig-mixed-models mailing list
> R-sig-mixed-models at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models
>
>
> End of R-sig-mixed-models Digest, Vol 34, Issue 19
> **************************************************
>   
-- 
Christine Adrion, Dipl.-Stat., MPH

LMU Ludwig-Maximilians-Universitaet Muenchen
IBE  - Institut fuer Medizinische Informations-
verarbeitung, Biometrie und Epidemiologie
Marchioninistr. 15
D- 81377 Muenchen
 
Tel.:      +49 (0)89 7095 - 7486
Fax:      +49 (0)89 7095 - 7491
eMail:    adrion at ibe.med.uni-muenchen.de
web:     http://www.ibe.med.uni-muenchen.de