FreeMat—Yet Another MATLAB Replacement
Many programs exist that try to serve as a replacement for MATLAB. They all differ in their capabilities—some extending beyond what is available in MATLAB, and others giving subsets of functions that focus on some problem area. In this article, let's look at another available option: FreeMat.
The main Web site for FreeMat is hosted on SourceForge. Installation for most Linux distributions should be as easy as using your friendly neighborhood package manager. FreeMat also is available as source code, as well as installation packages for Windows and Mac OS X. Once it's installed, you can go ahead and start it up. This brings up the main window with a working console that should be displaying the license information for FreeMat (Figure 1).
Figure 1. When you first start up FreeMat, you are given a fresh console.
As with most programs like FreeMat, you can do arithmetic right away. All of the arithmetic operations are overloaded to do "the right thing" automatically, based on the data types of the operands. The core data types include integers (8, 16 and 32 bits, both signed and unsigned), floating-point numbers (32 and 64 bits) and complex numbers (64 and 128 bits). Along with those, there is a core data structure, implemented as an N-dimensional array. The default value for N is 6, but is arbitrary. Also, FreeMat supports heterogeneous arrays, where different elements are actually different data types.
Commands in FreeMat are line-based. This means that a command is finished and executed when you press the Enter key. As soon as you do, the command is run, and the output is displayed immediately within your console. You should notice that results that are not saved to a variable are saved automatically to a temporary variable named "ans". You can use this temporary variable to access and reuse the results from previous commands. For example, if you want to find the volume of a cube of side length 3, you could so with:
--> 3 * 3 ans = 9 --> ans * 3 ans = 27
Of course, a much faster way would be to do something like this:
--> 3 * 3 * 3 ans = 27
Or, you could do this:
--> 3^3 ans = 27
If you don't want to clutter up your console with the output from intermediate calculations, you can tell FreeMat to hide that output by adding a semicolon to the end of the line. So, instead of this:
--> sin(10) ans = -0.5440
you would get this:
--> sin(10); -->
Joey Bernard has a background in both physics and computer science. This serves him well in his day job as a computational research consultant at the University of New Brunswick. He also teaches computational physics and parallel programming.