Monday, June 9, 2014

What would be good complements for Python and R?

No title
What would be good complements for Python and R?

by Christian Bolton

June 9, 2014

1

During 2014, I have been trying to expand and improve my computer programming skills, because of the long term benefits. I am not willing to do any more coursework or degree programs. I have already done far too much coursework, homework and exams.



Python and R are the 2 programming languages that I am convinced I should know how to use. I have already made some progress becoming familiar with Python and R.



My question is: which other programming languages would be a good complement for me to learn in the future?



There might be some objectives and contexts for which neither R nor Python is optimal. I certainly to do not know enough about computer programming to know which tool is optimal for which context and objective.



Would it be best to learn a programming language invented within the past few years, like Go?



Would it be best to learn the latest form of a programming language with a very long history, like Fortran 2008 or Common Lisp?



Would it be best to learn C only, or C++ only, or both C and C++?



My goal is to expand my long term raw computational power, in a way that is outside of academia and outside of all PhD programs and advanced degree programs.



I would especially like to be able to implement as many theories from statistics and numerical methods as possible. I do not know which theories from statistics and numerical methods I might find applicable in the future.



In case you want to know something about my academic background: I have done about 24 college level math classes, and zero college computer programming classes.

2 comments:

  1. For academia learn python and learn Scala. It runs on the JVM, where a lot of parallel computing is done. Skip R.

    ReplyDelete
  2. Julia. High level but much better performance than Python or R. A very neat multiple-dispatch type system. Metaprogramming for when you want to do code generation. Smaller community and fewer libraries than either Python or R since it's much newer, but also much easier to contribute to the language, standard library, and packages since so much of the language implementation and high-performance libraries are written in high-level Julia. This is opposed to the internals of Python or R which are in C and Fortran (for parts of NumPy, SciPy, and R), so far less accessible.

    ReplyDelete