Software
Chrest Shell
The Chrest Shell is a graphical application and library for running a variety of Chrest models and experiments. Currently, the shell supports a limited range of domain-types, but more functionality is available using the scripting interface. Detailed examples are provided for Lisp (using ABCL) and Ruby (using jruby), and illustrative examples given for groovy (a Java-like language), and clojure (a Lisp-like language). Refer to the documentation within the download package for more information.
Requirements: Java 6.0. To develop your own models, you will need either a Java compiler or a scripting language for the Java platform.
Download: jchrest-0.16.zip. Unpack the zip file and double-click on the 'jar' file; more information is contained in the file 'README.txt'. (Last update: 30/1/12. MD5SUM 6ca9a9826c0fa7624947fa1e0c4b9ae9)
Source code: can be obtained from Launchpad. License: GPL v 3.0.
Domains
The following domains are currently supported:
- verbal learning
- classification
- pattern recognition and recall
- visual attention and memory
Screenshots
| Model Training | Stimulus-Response Experiment | Model View |
|---|---|---|
|
|
|
Lisp Environment for Creating Models
For developing Chrest models in Lisp, we recommend the J editor and ABCL, a Java-based implementation of Lisp. All necessary files and startup scripts are contained in the following:
- jchrest-and-lisp.zip Unpack the zip file, and run the .sh or .bat file appropriate for your operating system; the file 'README.txt' contains further information. Last update: 8/5/10.
| Editing and running a Chrest model within J's lisp environment. |
|---|
|
CHREST++
The CHREST++ code is a Lisp implementation of CHREST, used mainly to develop models of chess expertise. The code is written in Lisp, and mostly tested using SBCL.
- Download chrest-code.tar.gz. (Last update: 5/2/10.)
Miscellaneous
Mini-Chrest is a version of CHREST with minimal features. It can be used for instructional purposes, and to become familiar with the implementation of the core learning algorithms.
- minichrest-lisp.tar.gz. (Last update: 10/9/10.)