An Extensible and Scalable
Agent-Based Simulation of Barter Economics
Mait Märdin and I (Pelle Evensen) did our master's thesis work in Computer Science replicating, re-implementing and extending Herbert Gintis BARTER-model.
"This thesis project studies a simulation of decentralised bilateral exchange economics, in which prices are private information and trading decisions are left to individual agents. We set to re-engineer the model devised by Herbert Gintis and take his Delphi version as the basis for providing a new, portable barter economics simulation tool in Java. By introducing some extension points for new agent and market behaviours, we provide simple means to implement variations on the original model. In particular, our system could be used to study the emergent properties of heterogeneous agent behaviours.
Through the addition of some default visualisation models we provide means for an improved intuitive understanding of the interaction between individual agents.
The multi-agent simulation library MASON is used as the underlying simulation platform. The results of running the software with various parameters are compared to the results from the original version to confirm the convergence of the two programs."
Since Mait was a student at Chalmers University of Technology and I was a student at Gothenburg University, the thesis is printed with two different covers. Same contents though.
| File | Contents |
|---|---|
| barterEconomy.tar.gz | Archive containing JAR-archives. Download this if you want to play around with the program. You will need Java SE Runtime Environment 6 or later. See Appendix A of the thesis for a short user's manual. |
| barterEconomySrc.tar.gz | Archive containing the source code for our program. If you are in an experimental mood and want to extend the program (or simply fix bugs), you will also need the MASON Multi-agent simulation toolkit. When we were developing our program we did find a few bugs in MASON that were fixed in the CVS version but not in MASON version 13 (or 14). |
| 200903.Evensen-Mardin-GU.pdf | Master's thesis, Gothenburg University version |
| 200903.Mardin-Evensen-CTH.pdf | Master's thesis, Chalmers University of Technology version |
| 2008-12-19-Evensen-Mardin-Presentation.pdf | Thesis presentation, not very self-contained. |
| Different convergence behaviour | We have some strong evidence that our program's convergence behaviour is different from Gintis' Delphi version. See Chapter 6 and Appendix D in our thesis. Note that this does not seem to be the case for all parameters. |
| Strange individual agent visualisation | When bringing up the individual agent visualisation window, some agent charts have the wrong size. We didn't have the time to figure out if this was a bug due to us or to MASON (or both). Figure A.7 from the thesis shows what it looks like when things work as they should. |
| Wrong colours in charts | When charting "Average Score", "Average Price" and "Producer Share" (Figures A.3--5 in the thesis), the colours are not the same as they are in the agent visualisation window. |
Please contact us if you use the program for any purpose or find it interesting in any way. Since we kind of doubt that anyone will, we have not (yet) found the energy to spend more time on the issues above.
We can be reached at barter@evensen.org.
It is unlikely that we will fix any of the above bugs or other ones you may encounter even though we are very interested in hearing about them. If you have patches that work, we will of course include them.
The thesis, source code and binaries are copyright Pelle Evensen, Mait Märdin 2009. The source is hereby released under the Perl Foundation's Artistic License 2.0.