R – the swiss army knife of the data scientist

R LogoThe video below, a talk by John D. Cook (via Flowingdata), is a very nice intro to R for the someone who wants to be a data scientist and have some notion or experience of programming. I have been beginning to look at R, but need a specific project to analyze in order to get into it. When learning a programming language (or any powerful tool, for that matter) it is important to get under the skin of it, to understand it to the point where you don’t look up the function or whatever in the manual because you intuitively know what it would be named, since you think like the developers. (I can’t claim any knowledge like that, except perhaps for IFPS (a defunct financial programming language), REXX (macro language for IBM mainframes), and Minitab (statistical package, rather marginalized now). Learning something to that level requires time and, most importantly, a need. We’ll see.

But it helps to have someone explain things, so I guess watching this video is not a waste of time. It wasn’t for me, anyway. And R certainly is the thing to learn, in this Big Data (whatever that may mean) world. (Though, as is said here, it was never designed for huge data sets. But huge data sets need models to work, and you build those on small data sets…)