***************************IMPORTANT NOTICE****************************
This version of Cubist is distributed under the licence agreement in
http://www.rulequest.com/eula.html.  Use of Cubist or of its documentation
signifies your willingness to be bound by this agreement.
***********************************************************************


The Cubist System
-----------------

This Cubist release contains the following components:

    *  a rule-based linear model generator ('cubist')
    *  a csh script ('xval') for multiple cross-validations
    *  an ancillary program ('summary') used by xval
    *  a grahical user interface ("gui") and an associated
       file ("gui.exe.so").

The grahical user interface requires Wine (http://winehq.org); Wine
is included in most recent Linux distributions.  Before using gui,
please run Cubist from the command-line interface to install your
Licence ID (see Installation below).

All programs are in directory 'bin'.  There is a html tutorial containing
instructions for using Cubist in the directory 'Doc'.

Installation
------------

Installation of Cubist requires a valid Licence ID.  This is a string of
hexadecimal digits containing spaces and terminated by a hash sign (#)
and may look something like this:

    9f23 4be2 a80f 0100#

The first time that you run Cubist, make sure that you are in the directory
that contains the cubist file.   Run Cubist from the command line, not from
the graphical user interface.  You will be prompted to enter your Licence
ID and, after it is checked, the cubist file will be overwritten.  You will
not be required to enter your Licence ID on subsequent uses of the
overwritten file, but you will need to do so if you attempt to run a copy
of the original file or if you obtain an updated version.

Examples
--------

The directory 'Data' contains some sample datasets.  Many interesting data
collections are maintained by the University of California, Irvine, in a
data repository.  For more information on how to access files from this
repository, send email to

	ml-repository@ics.uci.edu

Problems
--------

If you discover any bugs in the system, please email quinlan@rulequest.com.
