[R] piper diagram
jim at bitwrit.com.au
Tue Jun 10 14:34:13 CEST 2008
Dylan Beaudette wrote:
> Is anyone on the list familiar with an R implementation of Piper Diagrams?
> I am thinking that two calls to triax.plot (plotrix) along with some kind of
> affine-transformed standard plot would do the trick. Not so sure about the
> final layout, or a nice generalized version for something like lattice.
You're correct about the lower two triangle plots, triax.plot would
handle these (although there would probably be a bit of hacking about to
get the two plots in the right place). I don't understand the upper
rhombus. As a third triangle plot I can get it, but the bottom half
seems to be either redundant or nonsensical (you can't have negative
proportions of components). If I have some time, I'll look up "piper
plot" and try to work out what it all means.
More information about the R-help