CompStat: making R packages

In my Computational Statistics working group yesterday, Yihui Xie (a graduate student in the department) gave a talk about his experiences creating R packages. Yihui maintain 7 R packages: animation, formatR, Rd2roxygen, R2SWF, MSG, iBUGS, and fun. With two more in the works: cranvas and knitr. From the talk, the ones that I was particularly interested in are

formatR which allows you to write sloppy R code and have formatR fix it up as well as convert = signs to -> (or vice versa),
Rd2roxygen which is a counterpart to roxygen and aims at easing the process of documenting R packages by allowing the developer to write one document that contains both the code as well as the help documentation. roxygen separates these into their .R and .Rd files while Rd2roxygen does the opposite, i.e. when you want to update something in your code
iBUGS which is aimed at easing the massive number of arguments required when using R2WinBUGS. I did give this one a shot and it appears to mainly be a GUI for setting these arguments which doesn't interest me as much.
knitr which is meant to be an alternative to Sweave for creating living documents.

Give 'em a shot.

03 November 2011