What programming languages do you use? | Coursera Community

What programming languages do you use?

  • 6 November 2018
  • 5 replies

Userlevel 5
Badge +5
I'm interested in what programming languages are being used the most right now. Of course it is best to know a variety of languages and use the best one for the task at hand.

Particularly what languages do you use for Data Science projects?

I've added a poll, but please do comment and discuss as well!

Personally, I use Python mostly right now. At university I mainly worked with MATLAB, and through my career have gained experience with a few others including R, C#, F#, Julia.

Which programming language do you use the most?

5 replies

Hello Liz!
I participated in a recent ZDnet article on the subject and I think it is important for students to know that language is not as important as the added value of the implemented process. And when you deliver this value it's important that it be consistency, fault tolerance, and support scalability.
Take a look at the full story:
Userlevel 3
Badge +4
I like Python. Although I have experience with PHP and JS. And also I studied several courses and tried JAVA and Kotlin, but they seemed too complicated ...
At university I studied Pascal, and at school Qbasic, but these languages are generally useless now.
Userlevel 3
Badge +1
@Liz: I've been in software development for 50 years (now retired) and languages come and go. If you;re into software development, keep active in several, and strive to learn a new and different culture language every year or two.

In simple terms today I like Python and JavaScript.
For longevity in marketable skills, C and C++ are essential base skills. I used C for over 30 years.
Learn Haskell, it'll improve all your other programming skills even if you never use it in a project.

One of my favorite interview techniques was to ask about a candidate's favorite language "What is wrong with it that makes XXX so very complicated and necessary?") For example what is wrong with class and object features that means that dependency injection is so heavily used, and so intricate to implement.

When I took a Haskell course there was a strategy that was very effective: write the simplest possible expression that can produce a value of the intended type. Simple is a really good feature.
Userlevel 1
Badge +1
For data science it is most important to understand and implement algorithms.
Python is one of the languages that is really self describing. And one of the major reasons why I use it for data science projects especially machine learning is that it is very light. Anyway, for all my projects that have complicated algorithms, I use OCTAVE for trying out algorithms(since I have all datascience codes stored there). Mainly by breaking it into simpler problems and then in end I convert it into desired language and join the simpler and shorter algorithms.
Hope this helps!
Userlevel 3
Badge +1
Hardi has a good point. Many important pieces can be, and should be, implemented in different languages, or in different styles with one language. At different project stages you are dealing with different issues: "Does this idea work", "Can we make that work quickly enough", "Can we make that work at scale". Sometimes we should implement something in different ways as a learning method for new languages, libraries, tools.