Like many users and producers of R packages, I have never had any formal training in computer science. I’ve come to to the conclusion that this is a serious omission in a professional researcher’s training. Computer scientists and professional hackers have learned a lot about effective, efficient programming over the last five decades and it’s past time academic researchers begin to learn from this experience.

To that end, here are three books which I think all users of R could benefit from reading:

• Practical Common Lisp by Peter Seibel. Free online version.
• Structure and Interpretation of Computer Programs (SICP) by Harold Abelson, Gerald Jay Sussman, and Julie Sussman. Free online version.
• On Lisp by Paul Graham. Free PDF version.

Yes, I realize all of these are books that use Lisp/Scheme, and I realize the structure of Lisp can be a bit off-putting ((at '(first))). But, unless you want to learn Lisp, you shouldn’t try to read these books with the goal of understanding ever nuance of the syntax; rather, you should read them with an eye toward understand how a programming language should work for and not against you. For instance, each of these books shows, in a step-by-step approach, how seemingly complex programs can be effectively broken down into simple procedures that anyone with some programming experience can understand. A great deal of emphasis in these books is the important concepts of abstraction and non-duplication. If you’re like me and spend any significant amount of time going through various R packages, you’ll know what I mean when I say that the R community could benefit from less duplication and more abstraction in contributed packages (not to mention more comments!).