MATLAB vs Python vs Maple

[FelipeVv]

Hello everyone. I'm an undergraduate student (third year) and I know these software/language are very famous and powerful. If you check the web, you see lots of articles in which they compare them but, they're confusing. Each of these has advantages over the others. And there's no much about these software applied to chemistry, which is my main interest. So I come here to ask you guys if you use any of these software and what for. Also, if you have tried more than one of them, which one would you consider the most useful in chemistry? I'm very interested in multivariable analysis, machine learning, manipulating spectra (second derivatives, vector normalization,...). Thank you in advance.

[Usmivka]

I suggest that if you are learning a new language from scratch, start with freeware. Matlab requires an annual subscription that may not always be covered by your university. Maple costs money and I don't know anyone who uses it, so I can't say what it might offer that is worth the cost. I use Matlab extensively, and appreciate a lot of things about it, but it can be convoluted sometimes, and I wish that I had more freeware in my toolkit. R is a freeware equivalent to Matlab that would be a smart choice to work with if you are focused on statistical analysis, but doesn't seem to me to work as smoothly for machine learning. R also is widely used by biologists and ecologists, which brings up the point that if there is a standard in your field, you should probably use it in order to be able to take advantage of the models and toolkits that others make available for that software. Python seems powerful and relatively simple to use (I have a friend that swears by it, and he is a much better coder than I am), I personally have not tried to learn it yet, but it's on my to do list. A major benefit of python is that it can produce lovely images and figures for publication--making Matlab and R figures aesthetically pleasing can be time consuming without third party software like Illustrator or Gimp.

Edited June 12, 2017 by Usmivka

[avflinsch]

Both R and Python have hundreds of user contributed packages/libraries. There is a very good chance that someone has already created one that does exactly what you need, or something close enough that can be easily adapted to your needs. It is a good idea to learn both, and there is nothing to stop you from from doing part of the analysis in one, exporting the data midstream, then continuing in the other (and switching back if needed).

Personally, I use a combination of SPSS, Matlab, R, Python and other languages as needed. Some things are easier to do in one platform, and some things are easier to do in another.