SAS is a proprietary (and expensive) data and statistics language. It is used heavily in business and government, as opposed to R, which is used heavily in academia. It's better at handling large data sets than R is, but R is a more traditional language that most find easier to learn. (Also, the SAS Institute has a virtual lock on publishing, and the books suck. I know of one good SAS book, and I've been programming in SAS for 15 years). R used to have an advantage in graphs and charts, but SAS's new(ish) SG graphics puts them ahead of R at this point (minimal ink people will disagree with me).
Python is, of course, an up and comer in the data analytics field. When I retire this summer, I'm going to put some time into learning pandas and such. Then if things don't work out and I need a job, I will have the trifecta of SAS, R, and Python. Although my R is a bit rusty. I don't like it's overly functional paradigm and it's rather poor implementation of OOP. Of course, SAS is such and odd language it doesn't have any OOP. I think of SAS like Lisp: it's such a completely different paradigm, and you have to think in that paradigm to really make good use of the language.