useR! 2008 is almost done already. The last session I attended was in the Audimax, the largest lecture room of the conference site. Focus of the majority of the presentations was on teaching R and of motivating people to use it. However, the most interesting presentation was on R Analytic Flow, a rather simple application aimed at providing a hierarchical overview of your complex analyses.
R Analytic Flow has something very simple, but very interesting to offer for especially the more complex projects. Earlier, I wrote a project called MORET to organize your analysis, and R Analytic Flow has something similar to offer. Basically, R Analytic Flow offers a hierarchical view of your analysis. Each node of the hierarchical tree contains a basic function with corresponding parameters, which are executable directly. Also, future versions will allow for caching, so that the entire workspace is saved in the corresponding node. In that way, complex and time-consuming calculations do not need to be repeated. This is something I will surely try (see: http://www.ef-prime.com) and will help me to keep the overview on complex projects.
Perhaps, R Analytic Flow might even help people to start using R-Project, which was more explicitly the purpose of the other presentations. I’ve seen Nicholas Lewin-Koh presenting how he implemented to R-Project the analyses that are quite standard and often performed on his medical department. To give students a better feel for statistics, Adrian Bowman wrote simulation modules for often used distributions and analyses. Richard Pugh from Mango Solutions gave a nice presentation on the experiences of his company with teaching R-Project to people. I totally agree with him that at first a firm basis should be created, before the more interesting capabilities should be taught.