I wrote an article titled “Introduction to SciPy Programming for C# Developers” in the March 2016 issue of Microsoft’s MSDN Magazine. See https://msdn.microsoft.com/en-us/magazine/mt683799.aspx .
I think of “data science” as a combination of classical statistics techniques such as linear regression, and machine learning techniques such as neural network classification. The three most common open source tools for data science are the python language with the SciPy (“Scientific Python”) library, the R language, and the SciLab (or the roughly equivalent Octave) programming language and environment.
In the article, I explain that Python doesn’t have a native array type. So, to write data science programs with Python, you need to install the NumPy library (“Numerical Python) which has an array type, then SciPy (which has advanced functions for data science).
Python and SciPy are based on the C language, so C# and Java developers can learn SciPy programming quickly. Most of my colleagues prefer using Python and SciPy to using R or SciLab.