Technology for Data Processing

Warning! Some information on this page is older than 6 years now. I keep it for reference, but it probably doesn't reflect my current knowledge and beliefs.

Mon
04
Oct 2010

When doing some engineering work on a computer, no matter if gamedev or any other field, there is sometimes a need to process or visualize some tabular data, especially numbers, e.g. statistics about performance or something gathered during program execution. What technology is best for this purpose? At the moment I know about following solutions:

Spreadsheet software, like Microsoft Excel or OpenOffice Calc. They can import CSV files and draw great variety of plots, but for more advanced data processing it would be useful to have something more like a programming language.

At the other end of the spectrum, we can write normal C++ or C# programs to do the work. It can be hard though as we have to code everything on our own, including data structures, loading files and drawing plots. It would be nice to use some higher-level, scripting language with rich standard library, built-in data structures and the like.

Any scripting language can do this. For example, PHP (which I know the best) can be used as a normal scripting language, not only in connection with a web server. All in all, its name means "PHP Hypertext Prerocessor", because it's suited for processing strings and text files.

A technology designed specially for the purpose of crunching numbers is Matlab and its free alternative - Scilab. It provides its own programming language with convenient built-in data types like matrix and is also able to draw plots.

Python is something in between - a normal scripting language with weird but nice syntax and a language-level support for operations like array slicing and complex numbers. It looks like many scientists and engineers use it for computation and data analysis, because there are Python libraries designed for this, like Numpy and Scipy.

There is also The R Project metioned at the Nick Darnells' Blog. It looks like another environment with its own, different programming language, designed especially for statistical computing. It can also load data from files and draw plots.

And finally, some tasks can be accomplished in environments that allow playing around with computer graphics, like EvalDraw (with its own, C-like simple programming language) or Processing (with a language based on Java).

So there are many possibilities in this subject. I've played a bit with all of them at some time, but obviously learning chosen one thoroughly would require much time and effort. So maybe you can help me decide? Which solution would you recommend? Personally I feel a little more convinced to Python, because it is a general purpose language that I can use in different fields too, like coding Blender plugins.

Comments | #math #languages #tools Share

Comments

[Download] [Dropbox] [pub] [Mirror] [Privacy policy]
Copyright © 2004-2024