[R] piper diagram

[Jim Lemon]

Dylan Beaudette wrote:
> Hi,
> 
> Is anyone on the list familiar with an R implementation of Piper Diagrams?
> 
> Example:
> http://faculty.uml.edu/nelson_eby/89.315/IMAGES/Figure%209-78.jpg
> 
> 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.
> 
Hi Dylan,
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.

Jim