Ox vs. Gauss

If you need a matrix programming language and are considering Gauss, think about using Ox instead. It was developed by Jurgen Doornik of Oxford and is better-designed, generally faster, portable across more platforms, and free for academic use.

The basic syntax of Ox is similar to Gauss, which makes it easy to switch. (However, Ox does not include the extra packages that can be purchased for Gauss, such as MAXLIK). A nice feature of Ox is that it can read data in a lot of formats: ASCII, Excel, Lotus, Gauss and Stata.

To download a Windows version for your own computer, or to find out anything else about Ox, see the author's home page.

Here's an example Ox program that reads a matrix from an Excel spreadsheet and computes its eigenvalues. Everything following a // is a comment and not a required part of the program:

// ---------------------------
// Computing eigenvalues in Ox
// ---------------------------

// Required Ox header information

#include 

// Main Ox routine

main()
{
   // Declare variables for the array and the eigenvalues

   decl a,eigs;

   // Read an Excel spreadsheet to load values into a  

   a = loadmat("x.xls");

   // Compute the eigenvalues and store them in eigs.  The & is
   // needed so that Ox can use eigs to return the eigenvalues.

   eigen( a, &eigs );

   // Print them out

   print( eigs );
}