Genetic Evolution through Combination of Objects


GECO is George P.W. Williams' CLOS-based toolkit for writing genetic algorithms. It is a fairly flexible framework that can be used for experimentation and for the solution of a variety of problems. Documentation is available, and it includes two simple examples.

Version 2.01a has been tested on SBCL and CMUCL.


October 25, 2011
Released version 2.01a. Made sure everything still works, and removed old emacs-related cruft.
April 2, 2006
Version 2.01 released. Changes: added an ASDF system definition file, and removed the minimal traces of bitrot. The examples work.
Nov. 29, 1993
Version 2.0 released.


A paper describing the framework can be found in the CMU AI Achive. The .ps file seems to be damaged, you might have to convert to a pdf file to see more than the first page. You can download it here.



Recommended: using ASDF.

  1. Download the source code and unpack it in the directory where you want to install it.
  2. Make a symbolic link for in your central registry to the .asd file of GECO.
    ln -s /my/central/registry/geco.asd /path/of/code/geco.asd
  3. Start your lisp implementation, and compile (and load) the code:
    (asdf:operate 'asdf:load-op 'geco)

A system definition file is also included.


Extensions, fixes, and additional examples are welcome. If you have patches, extensions, ideas, etc. just drop by at the geco-devel mailing list and tell us about it.

Development is tracked using darcs. The latest version of GECO is in the darcs repository at:

$ darcs get


If you have problems running GECO, or have found a bug, consider telling us at the geco-devel mailing list.