R help - Mar 2004 - lme question

Hi,

I have a dataset like this,
> testdata
Grouped Data: expr ~ visit | subject
      expr visit subject
1 6.502782    V1       A
2 6.354506    V1       B
3 6.349184    V1       C
4 6.386301    V2       A
5 6.376405    V2       B
6 6.758640    V2       C
7 6.414142    V3       A
8 6.354521    V3       B
9 6.396636    V3       C

I tried the command
> testresult=lme(expr~visit,data=testdata,random=~visit|subject)
I got the following error message.
Error in MEestimate(lmeSt, grps) : Singularity in backsolve at level 0, block 1
In addition: There were 50 or more warnings (use warnings() to see the
first 50)

Could someone give me some hint on what went wrong?  Thanks.

Donglei Hu

Hi Donglei,

what is your goal in fitting this model?  The statement that you've used is 
creating an overparameterized model, as it fits a random intercept for each 
visit within each subject.  I wonder if you might prefer

testresult <- lme(expr~visit, data=testdata, random=~1|subject)

which will fit a random intercept for each subject?

Andrew


On Monday 22 March 2004 14:29, donghu at itsa.ucsf.edu
wrote:
> Hi,
>
> I have a dataset like this,
>
> > testdata
>
> Grouped Data: expr ~ visit | subject
>       expr visit subject
> 1 6.502782    V1       A
> 2 6.354506    V1       B
> 3 6.349184    V1       C
> 4 6.386301    V2       A
> 5 6.376405    V2       B
> 6 6.758640    V2       C
> 7 6.414142    V3       A
> 8 6.354521    V3       B
> 9 6.396636    V3       C
>
> I tried the command
>
> > testresult=lme(expr~visit,data=testdata,random=~visit|subject)
>
> I got the following error message.
> Error in MEestimate(lmeSt, grps) : Singularity in backsolve at level 0,
> block 1 In addition: There were 50 or more warnings (use warnings() to see
> the first 50)
>
> Could someone give me some hint on what went wrong?  Thanks.
>
> Donglei Hu
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://www.stat.math.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide!
> http://www.R-project.org/posting-guide.html
-- 
Andrew Robinson                      Ph: 208 885 7115
Department of Forest Resources       Fa: 208 885 6226
University of Idaho                  E : andrewr at uidaho.edu
PO Box 441133                        W : http://www.uidaho.edu/~andrewr
Moscow ID 83843                      Or: http://www.biometrics.uidaho.edu
No statement above necessarily represents my employer's opinion.

Dear All,
I need help both with analytical and computational aspects.

My problem: I have counted and measured increments on otoliths (inner-ear bony 
structures) of fish larvae collected at different times of day. This was 
repeated three times for each fish larvae (in random order and no particular 
dates). Then, I calculated the ratio of marginal-to-previous increment for each 
fish and reading. I wish to study if differences exist in ratios between hours 
of day (accounting for differences in readings within each fish) - in 
data.frame format... 

code   hour   read.no   ratio
33001  5      1         0.56
33003  7      1         0.93
37123  19     1         0.23
...
33001  5      2         0.62
33003  7      2         0.87
37123  19     2         0.22
...
37123  19     3         0.26
...

My solution: Use function lme, with hr<-factor(hour)...
> hr.lme<-lme(sqrt(ratio)~hr,random=~1|code/read.no)
> anova(hr.lme) ; summary (hr.lme)
Does this commands correspond to the initial problem statement? 

Thanks in advance.
Regards, Eduardo Esteves