DMtools
=======

A toolbox for efficient and flexible data mining.

Authors:                          Contact and more information:
 - Ole M. Nielsen                 - URL:    http://csl.anu.edu.au/ml/dm/
 - Peter Christen                 - E-Mail: Ole.Nielsen@anu.edu.au
 - Markus Hegland                           Peter.Christen@anu.edu.au
 - Timothy Hancock
 - Tatiana Semenova

Introduction
------------

The DMtools are developed by the Data Mining group at the Australian National
University (ANU), Canberra. Efficient and flexible data exploration and
analysis are often crucial for the successful outcome of a data mining project.
The DMtools help the data miner with efficient handling of common tasks in
data mining.

Using the scripting language Python (www.python.org) the DMtools are based on
two core features: (1) Caching of time consuming functions including, but not
limited to, database queries, and (2) parallelism within a collection of
independent queries. The toolbox provides a number of routines for basic data
mining tasks on top of which the user can add more functions - mainly domain
and data collection dependent - for complex data mining tasks.

For more details see our our web page at

  http://csl.anu.edu.au/ml/dm/

and a scientific publication describing the toolbox

  A Toolbox Approach to Flexible and Efficient Data Mining,

to be presented at the 9th International Python Conference (Long Beach, March
2001) and the 5th Pacific-Asia Conference on Knowledge Discovery and Data
Mining (PAKDD, Hong Kong, April 2001), available online at

  http://csl.anu.edu.au/ml/dm/dm_publications.html

Files
-----
This distribution of the DMtools contains the following files:
- caching.py       The caching module
- database.py      The database access module
- aggregation.py   A module providing several aggregation functions
- README           This readme file
- LICENSE          A copy of the GNU General Public License

Installation
------------
- To be able to use the DMtools, you need Python installed on your system.
  See the Python homepage (http://www.python.org) for downloading and
  installing.

- You also need a relational database for which a Python database module is
  available. For more information about database access within Python have a
  look at the Python database SIG: http://www.python.org/topics/database/
  We are using MySQL, available from: http://www.mysql.com

- We use the MySQLdb.py module by Andy Dustman, see:
  http://dustman.net/andy/python/MySQLdb/
  The DMtools are tested with version 0.2.2 of MySQLdb.

- The caching.py module works optimally if the zlib module 
  (http://www.cdrom.com/pub/infozip/zlib/) is installed.

- Both the caching.py and database.py modules have a dictionary of options at
  the beginning of the code. Please change the entries to your appropriate
  values, using the function set_option() in either module.

- The caching module contains a built-in self-test function which can be
  started with:
  >>> import caching
  >>> caching.test()

License
-------
This software is distributed under the GNU General Public License. For more
details see the LICENSE file.