I was wondering about the experiences using various solutions to math/statistics solution to use in a Processing app. Namely, I used mainly Jama and Apache Commons math for simple matrix manipulation but in the long run it becomes tedious to write code for i.e. discriminant analysis. So I came up with two possibilities, because I mainly work with R: 

1. JRI/Rengine/rJava solutions 

2. Clojure/Incanter

As for number one solution it seems to work when including Rengine .jars in the sketch folder but for that R and Rserver must be working. Also there are sometimes path-related problems and plus I`n not sure is is possible to even publish working app as standalone. Does anyone knows is there a possibility of Processing reading data structure that is currently hanging in the R workspace?

As for Clojure/Incanter it seemed to me very interesting because R-like syntax but even as incanter.processing library exists I don`t know how to invoke Incanter from Processing or is that a possibility at all?