I am working on learning R for a predictive analytics project. I have been using T-SQL for a long time and all of my data is in SQL. Anything that's not in SQL, I import to SQL and work with it from there (I want everything stored in the data warehouse for safekeeping anyway).

I'm doing the Coursera Data Science track, and I just feel like I might be wasting my time learning a lot of this stuff. I can do the data shaping and exploration in SQL, so why learn how to do it in R? Will there be a benefit in the long run to learning it in R?

I am planning to use R to build my predictive model, but it seems like I could generate a nice clean csv file in SQL and then just use R from that point forward.

Any thoughts?