The program Structural VAR was written to provide a simple Graphical User Interface (GUI) for estimating vector autoregressive (VAR) models using Maximum Likelihood (classical approach) and Markov Chain Monte Carlo (MCMC) methods (Bayesian approach). At the time when this project was started (late 2001) no such software seemed to exist. At least not in the public domain and to my own knowledge. Today, this is no longer the case, but I think there's a place in the software universe for SVAR, as I tend to call it myself. For those of you interested in languages, SVAR is the word for answer in Swedish (and a few other languages as well).
SVAR comes in two flavors. It can either be run under MatLab, or compiled (with the MatLab compiler with the C/C++ Math and Graphics libraries) and executed without MatLab.
Moreover, under Windoze, it can be installed (and uninstalled) with the setup program that's available for download from the homepage (see the About page). The setup program adds shortcuts to your Start|Programs menu (unless you don't want this), which gives you easy access to SVAR and its documentation. It also adds a shortcut to the software's homepage. If you so choose, the setup program can add shortcuts to SVAR on your desktop and on the Quick Launch toolbar. Moreover, it can associate *.var files, which contain information about your VAR model, with SVAR so that you can easily run the program by (double) clicking on a *.var file on your desktop or in Windoze Explorer. In addition, the setup program can associate *.fig files with SVAR. Unless you're going to experiment with the code or compile it (or use some other platform) the setup program is the simplest way to install the software and start putting it to use.
When running the software via MatLab, cd to the directory where the m-files are stored and then write var at the MatLab prompt. For the stand-alone release you just click on svar.exe (or var.exe if you insist on having a DOS style window) with your mouse or cd to the directory where svar.exe is located and write svar at the command line prompt. You can also run SVAR with Command Line Switches.
If you discover any bugs, strange behavior, or have any suggestions, please contact me at the email address given in the About section.