[R] PCA, Source analysis and Unmixing, environmental forensics

Gabor Grothendieck ggrothendieck at gmail.com
Thu Mar 23 22:01:07 CET 2006


Review all the task views at:

http://cran.r-project.org/src/contrib/Views/



On 3/23/06, Mike Bock <mbock at environcorp.com> wrote:
>
> I am using R for environmental forensics (determination of the sources
> and/or groupings in mixtures of organic chemicals in the field). The
> goal is to determine in there are groups of samples with
> similar/dissimilar compositions, and to assign samples to a potential
> source or a mixture of sources based on the composition (unmixing and
> source allocation). Typically there are 10 to 50 chemicals that have
> been analyzed in all of the samples. In most cases concentrations are
> converted to proportion of total as we are interested in composition
> rather that simple dilution.
>
> I have had great success with ratio analysis; simple exploratory
> analysis such a property property plots etc; and cluster analysis such
> as principal components analysis (PCA) and hierarchal cluster analysis.
> I have also been experimenting with glyph representation, k-means
> clusters, and similar procedures as documented in MASS. More recently I
> have been experimenting with Independent Components Analysis (ICA).
> Another commonly used procedure is polytopic vector analysis (PVA), a
> procedure for which I have no implementation in R sand so I have not
> tried it out. My question is can anyone suggest :
> A) Other procures in R that I may have missed and should investigate
> B) Your experience and/or hints for using ICA and presenting the results
> C) An implementation for PVA in R I have not found
>
> Thanks in advance,
>
> Mike
>
>
>
>
>        [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html
>




More information about the R-help mailing list