The Oxford English Dictionary includes more than 600,000 words, yet most of us get by in our day-to-day lives with a vocabulary of just a few thousand. In a similar vein, the R language includes thousands of functions: when you start up R 2.13, you have 2832 functions at your disposal:
This includes only the base packages attached by default: add in the recommended packages and all the packages on CRAN and you're easily into five figures.
But like English, you can make do with a much smaller vocabulary of R functions and still express yourself quite effectively. Hadley Wickham has published his essential vocabulary of R functions, listing just those functions you need for basic object operations, statistical analysis, working with the R system, data and text I/O, and dealing with specialized data types like dates and strings. There are about 350 functions on the list. This is a great idea: like many open-source projects, R has amassed many new functions (and indeed packages) that supersede (but not deprecate) older functionality, and it helps new users especially just to focus on those functions that are best for the job. Developing a small vocabulary of essential R functions is an important step forward.
github (Hadley Wickham): Vocabulary
Definitely some essential commands in the list, but also a fair number that are rather unusual. Hyperbolic trig functions, and "birthday" for example, are hard to imagine as "essential".
Posted by: Kevin Wright | May 11, 2011 at 20:17
@Kevin: Agreed that there are some oddities, but the page is a wiki, so you can always edit it if you feel strongly about something.
I've just added a few words that I thought were missing (repeat, nlevels, class, is and as).
Posted by: Richie | May 12, 2011 at 05:47
These kinds of posts are really helpful. As R is so expansive, I generally get 'stuck' into ways of doing things that, if I had known better, could be done very simply with pre-existing functions.
Great List!
Posted by: Kevin Emerson | May 13, 2011 at 10:00