The evolution of P/F vs T/K raises important questions

[npcompl33t]

In particular, the split of Herbal-A into Golf 1a, 1b, and 1.5a seems unjustified. 

I thought we at least agreed these had different bigram statistics? From the MDS plot. I don't think the variation between them is random noise, the text is different between these. 1b has almost half the numbere of chols as 1a -- its substantially different. The statistical criteraon to split them was they seemed like distinct enough populations to warrant it. With enough in-group similarity. I don't think it distorts anything for this analysis to split them that way.