At the UNIX prompt, type "octave", to run the program. You should then see an "octave>" prompt.
If you do not find Octave on your computer, e.g. if you obtain the message, "octave: Command not found." you should contact your system administrator, and request that the Octave program be installed on your computer. Octave is Free Source. There is no cost or license fee needed for Octave, and you can get the program source code.
Therefore there is no reason why Octave should not be installed on every computer in existence, or at least on every computer you are likely to encounter.
If you are the system adminstrator of your computer, you can obtain Octave, as a Red Hat "rpm" file if you are running Red Hat Linux.
If you are running Debian GNU/Linux, it is even easier to install. All you need to do is type, from a root shell, "apt-get install octave".
Matlab is equivalent, to Octave, but is not as good as Octave in many ways. Matlab is also deficient in some ways. Moreover, Matlab requires a license or a debugger to get it to work. This creates inconveniences for installation and maintenance of the program.
Moreover, if you are going to invest the time to learn one or the other, it is strongly advised that you choose to learn Octave rather than Matlab, because you will always have the program source code for Octave which you can compile on any new computer architecture you might find yourself using, whereas with Matlab, all you will have is an executable program (with built in sabotage to make it not work unless you put in various license keys, and the like).
A major reason motivating our decision to recommend Octave rather than Matlab is the ease with which students can freely obtain Octave, so that they can work at home on their home computers.
Moreover, the Free Source spirit of Octave follows the academic tradition of disclosure and publication of scientific results, even when those results are in the domain of computer science. Because Octave is developed by a scientific community through freely published innovations, it has become far better than Matlab's collection of unpublished trade secrets.
However, if you feel strongly that you want to use Matlab, instead of Octave, you should still be able to do these labs in Matlab, perhaps with occasional minor changes to the Octave scripts provided.
--S. Mann, 2001