Open VS Closed Analysis Languages

(This article was first published on Deciphering life: One bit at a time :: R, and kindly contributed to R-bloggers)

Open VS Closed Analysis Languages


I think data scientists should choose to learn open languages such as R and python because they are open in the sense that anyone can obtain them, use them and modify them for free, and this has lead to large, robust groups of users, making it more likely that packages exist that you can use, and others can easily build on your own work.

Why the debate?

This was sparked by a comment on twitter suggesting that data scientists and analysts need to be polyglots, that they should know more than one programming language or analysis framework (the full conversation of tweets can be found here)

The commenter suggested knowing at least two of:

  • R
  • Python
  • Matlab
  • SPSS
  • SAS
  • Julia
  • Octave

My comment back was that one should really evaluate whether SPSS, SAS and Matlab should be on this list, as they are closed languages, not open, or free.

I want to expand on why I made that comment. Let me be forthright, I have not used SPSS, nor SAS, but I have programmed in MatLab and R extensively, and dabbled in Python.

I also think it is a good thing for data scientists to know more than one language. Just to be clear, I am NOT arguing that point.


What is a closed analysis language? I would say that there are three types of closed languages:

  1. those that are not free (but still may their source code openly available)
  2. those where the underlying engine is closed source
  3. those where one cannot write their own functions to expand on what is already available

Now, MatLab fits the first two categories. It is rather expensive to get a license for, the license can be restrictive (I know they have had a lot of abuse of licenses in the past, and they are trying to avoid that), and you are not expected to poke around in the internals of the MatLab engine. Oh, and if you want more than the base engine, expect to pay heavily for add-on packages.

However, it is possible to write add-on's for MatLab. I have previously written a few.

Closed problem: Checking results

So why are closed languages a problem? A closed language that does not make it possible to examine the underlying functionality of the analysis engine has two problems:

  1. surety that calculations are done correctly
  2. the ability of others to run and check results

Both of these are issues that are really important in science. I would consider a data scientist to be doing actual science, so others should be able to scrutinize their work. The best scrutiny, is for others to be able to actually run their code. If I can't run your code, then how do I know what you did is right?? If I can't afford a copy of MatLab to run your code (assuming you made it available, you did provide the source for the analysis, right?), that is a bad thing.

Closed problem: Re-using code

Of course, the other problem is that with a closed language you have made it impossible for others to easily make use of your analysis. Sure, they could code it up in another language, but unless it is the be-all and end-all of analysis methods, I'm not going to bother. I don't have a license for MatLab, or SPSS, or SAS, and I can't afford it; therefore I'm not going to use your method / code nor give you a citation or credit.

Solution: Open languages

Languages like R and python, they don't have these problems. If I wonder how a function in the base distribution of R or python works, I can go look at the source. If I find a bug, I can suggest a fix, or fix it myself and tell others about it. In addition, if I write code to do an analysis, I can make it available and know that others should have the ability to examine it, including re-running it, in addition to using it for themselves, if it is licensed appropriately. This is the way science should work.

What should you use?

Some would argue that you should use MatLab, SAS, and SPSS because they have been around for a while, and are the standard. I would argue that you should not use them because they are controlled by single corporate entities, who are only interested in what will get people to buy their product and use it. You should use software that others are using, and that others will be able to use, regardless of income.

R is being used in lots of different places, by lots of different people for statistics, bioinformatics, visualization, and as a general functional language. Python is a great general purpose language that provides a lot of functional glue for doing lots of different things.

To leave a comment for the author, please follow the link and comment on their blog: Deciphering life: One bit at a time :: R. offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...

If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.


Mango solutions

RStudio homepage

Zero Inflated Models and Generalized Linear Mixed Models with R

Quantide: statistical consulting and training


CRC R books series

Contact us if you wish to help support R-bloggers, and place your banner here.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)