[R] repeatability/intraclass with nested levels

[Bert Gunter]

This is an example of what is commonly known as "functional data analysis,"
or "fda" for short. You can search on this term to learn more. R has
an fda package that deals with this sort of data -- and very likely
others besides: essentially you have a 200-d response vector and so
require a multivariate analysis that deals with the inherent
correlation among the wavelengths. fda seems the appropriate way to
handle this to me, but there are certainly other approaches, depending
on your context.

However, due to your relative ignorance of these matters, you would do
well to consult a local statistician to help get you started and guide
your initial forays into this arena.

-- Bert

On Tue, Oct 12, 2010 at 7:32 AM, Nevil Amos <nevil.amos at gmail.com> wrote:
>  I have a spectrophotometric dataset with repeated measures of a value
> at 200 wavelengths for each of  150 individuals.
>
> I would like to use the repeated samples to at each wavelength to look
> at measurement/observer error, compared to difference between
> individuals error
>
> I have looked at doing this with icc{irr} or using an anova approach,
> but I am unclear how to acheive this given that there is a nested
> structure- the value varies between wavelengths within individuals.
>
> Many thanks
>
> Nevil Amos
>
> data is structured thus:
>
>>       ANWC_NO Wavelength Repeat    value
>> 1       00239        300       r1 0.079501
>> 2       00239        302       r1 0.084113
>> 3       00239        304       r1 0.087697
>> 202     11157        300       r1 0.008449
>> 203     11157        302       r1 0.009489
>> 204     11157        304       r1 0.010142
>> 403     11158        300       r1 0.026999
>> 404     11158        302       r1 0.029612
>> 405     11158        304       r1 0.030271
>> 604     11159        300       r1 0.032784
>> 66934   38711        300       r3 0.062081
>> 66935   38711        302       r3 0.065137
>> 66936   38711        304       r3 0.067092
>> 67135   38712        300       r3 0.029389
>> 67136   38712        302       r3 0.030192
>> 67137   38712        304       r3 0.030860
>> 67336   38993        300       r3 0.045720
>> 67337   38993        302       r3 0.048060
>> 67338   38993        304       r3 0.051898
>> 67537   38994        300       r3 0.051815
>> 67538   38994        302       r3 0.055162
>
>
>        [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> 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.
>



-- 
Bert Gunter
Genentech Nonclinical Biostatistics