I have some familiarity with both so it’s hard to decide. Also, is there an IDE for python that is equivalent as RStudio for R? I guess Anaconda or Canopy are not ideal enough.
For the beginner, I would recommend you to learn Python. Relatively, R is a bit esoteric in syntax, so it is hard for you to be well-versed. What’s more, the performance of R is really bad. The amount of data to be analyzed by a data scientist is never to be small, but for R, it runs to traverse a file at a speed of more than 10 times slower than Python. You will be unable to use it when the data size grows large, obviously, you could not just deal with in-memory data (even if in-memory loop is done, Python also runs several times faster than R, unless the library functions of R can be used only). R interpreter looks as if it is not developed by computer professionals.
But, if professional mathematical and statistical calculations are involved, there is nothing for choice but R. Python’s statistical package is still a little more primitive, and developed as an initial tool with far lower maturity than R. The defect of Python is that these types of professional data object and function library are not rich enough.
In fact, what the data scientists more often do is conventional structured data analytics with relatively simpler basic operation but with rich combinations and flexible language. In this case, esProc is preferred, for structured operation, esProc provides a Table Sequence which is more powerful than Data Frame of R, also supports for cursor-style file computing, presenting a performance equivalent to Python; in particular, multi-threaded parallel computing process can be programmed much easier in esProc than in Python (R almost does not support multithread).