Statisticians tend to learn R in school. It's quirky structures are a major source of prototyping slowdown/technical debt. I recommend using Python/Pydata when people ask my advice in starting out. Once one learns Python then R isn't too bad to pick up.