This is an attempt by Tim Triche, Jr. [email protected] to interface Structurama with R via Rcpp. The logic is pretty simple, I have a bunch of patients where I think there's population stratification and I would like to fit a model with an unknown number of ancestor populations to try and shake out the fuzz. Basically I don't care for the "first 10 PCs" approach and I recognize that everyone loves Structure because of its graphics, plus I like Dirichlet process models. So why not use my problem (AML background mutation rates) to motivate a solution. If I can avoid leaving R, I can Sweave the supplementary methods part of the paper.

Thus was born rStructurama: like Structurama, but without the formatting bits.

Rest assured that any bugs are highly likely to have arisen from my coding, not Huelsenbeck's, but if you want to fork the repo and have a go at it, please do. Otherwise, you can get a clean copy of the C++ source from Huelsenbeck's site.

In no event should Huelsenbeck be blamed for my shitty code, however.