[R] Polytopic Vector Analysis (PVA)

[Kjetil Brinchmann Halvorsen]

Melanie Edwards wrote:
> I am curious whether anybody has or is developing this capability within
> R.  I have not found any software yet that has this capability and I am
> not sure whether it is too new a method and nobody is actually using it,

googling around I found that this (PVA) is a variant of factor analysis
restricted to find non-negative factors. It is not a new method, 
although maybe the name is. This has been/is used for instance in
air quality monitoring to identify sources of pollution, and if you have
some prior information about possible sources that maybe could be used 
to. I guess this could be called 'source unmixing' or something similar,
which indicatyes a similarity with independent component analysis. Maybe 
enough to restrict optimization to non-negative values?

help.search("factor analysis")  shows that factor analysis is multiply 
implemented for R, so maybe there is something, and if not, maybe
simple to adapt.

Kjetil Halvorsen

> or if there are other means to get the same analysis that I do not know
> of.  Any information regarding developments or use of this method would
> be helpful.
> 
> Melanie Edwards
> 
> 
> 
> 	[[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
>