IODA for NetLogo

Authors

Sébastien PICAULT
Nathan HALLEZ
Philippe MATHIEU

Download and install

The IODA NetLogo package can be downloaded from this page, either as a zip file or through git.

To install the IODA NetLogo extension for use by any model, place the ioda directory either in the extensions directory in the NetLogo directory, or in the same folder as the model that uses it.

For more information about other ways of installing an extension, please refer to the NetLogo Extensions Guide.

In order to use the IODA simulation engine, your NetLogo code must include the file named IODA_3_0.nls. Thus, if the IODA folder is installed in the NetLogo directory, it could be useful to put a copy of this file (or a symbolic link) next to your own NetLogo model, in order to use a relative path instead of an absolute path.

Getting started

See the Tutorial for a comprehensive, step-by-step guide of using the IODA methodology within NetLogo simulations. You can start with Tutorial #0 ("Template"), copy the directory, and adapt it to your needs.

Open any folder in the tutorials directory, double-click the model.nlogo file, read the "Info" tab, test features in the interface, look at the code !...

What is it?

This software is an extension for NetLogo (Wilensky, 1999), version 6 or above (tested with NetLogo 6.4).

The IODA NetLogo extension provides NetLogo users with many features of the IODA Methodology – "IODA" stands for "Interaction-Oriented Design of Agent simulations" (Kubera et al., 2008, 2011).

The IODA approach aims at simplifying the design and reusability of individual-based or agent-based simulations. It relies upon the following key ideas:

The declarative parts of the model should be separated from the procedural parts as far as possible.
- declarative parts include the nature of the agents and how they can interact
- procedural parts include essentially the simulation engine (scheduling, action selection algorithms…)
Thus, any relevant entity of a model should be represented by an agent (Kubera et al., 2010).
Any behavior is encoded through a generic rule called an interaction. Basically, an interaction is a condition/action rule involving two agents: a source that performs the interaction and a target which undergoes it. Interactions must be designed independently from the specificities of agents (and vice-versa).
The matching between agents and interactions is ensured through:
- an interaction matrix which indicates what interactions an agent may perform or undergo, and with what other agents
- a generic simulation engine which asks potential source agents to select and perform an interaction and a target agent.

Example of two agents and the interactions they can perform or undergo

Thus the design of an individual-based model is split into three main tasks, through a top-down process:

write the interaction matrix: who can do what with whom ?
describe the interactions i.e. each behavioral rule
write the elementary capabilities of the agents i.e. small code primitives

The goal of this extension is to allow NetLogo users to get familiar with the IODA approach, and use it within their models, for research or educational purpose.

For more information on the IODA approach, please consult the IODA NetLogo Tutorials, the IODA webpage and the research papers mentioned below.

Developing an interaction-oriented model with IODA NetLogo

The main usage pattern of IODA in NetLogo models is as follows: The user creates a model for the problem to study, according to the IODA methodology (Kubera et al. 2011). A IODA model is composed of the following elements:

A list of agents: those agents are supported by NetLogo turtles (or user-defined breeds) and patches; they can be endowed with their own variables and reporters/commands;
A list of interactions: an interaction is a condition/actions rule, which captures the abstract features of a behavior; it involves source agent (which perform the interaction) and one (or more) target agents (which undergo the interaction). The definition of interactions can be done manually in the NetLogo program, but you are recommended to describe them in a text file, like in the tutorial examples.
An interaction matrix and an update matrix which specify how interactions are assigned to agents. The update matrix contains interactions that each agent tries to perform in the beginning of a simulation step. The interaction matrix specifies what interactions an agent X may perform over an agent Y. Here again, you are recommended to write those matrices in a text file (CSV format), like in the examples.
A list of agent primitives: these primitives are reporters and commands that will be evaluated or run by the agents when trying to perform or undergo interactions. Interactions are written using abstract primitives: thus, agents must implement concrete primitives according to the interactions they can perform or undergo. For instance, if agents of the breed wolves can perform (i.e. be the source) of an interaction that contains the abstract primitives hungry? (in trigger/condition) and follow (in actions), then two concrete primitives must be written: a reporter named wolves::hungry?, and a command named wolves::follow.

The IODA NetLogo extension provides tools for checking the consistency of a model, i.e. to ensure that all concrete primitives have been written with respect to the definition of interactions, the interaction matrix and the update matrix.

After defining those items, you have to write a setup procedure to initialize your simulation and a go procedure to run it. If you simply want to make use of the predefined IODA simulation engine, just put ioda:setup at the end of your setup procedure, and ioda:go at the beginning of your go procedure. That's all!

References

Y. KUBERA, P. MATHIEU and S. PICAULT (2011), "IODA: an interaction-oriented approach for multi-agent based simulations", in: Journal of Autonomous Agents and Multi-Agent Systems (JAAMAS), vol. 23 (3), p. 303-343, Springer DOI: 10.1007/s10458-010-9164-z.

Y. KUBERA, P. MATHIEU and S. PICAULT (2010), "Everything can be Agent!", in: van der Hoek et al. (eds.), Proceedings of the 9th International Joint Conference on Autonomous Agents and Multi-Agent Systems (AAMAS 2010), Toronto, p. 1547-1548.

Y. KUBERA, P. MATHIEU and S. PICAULT (2008), "Interaction-Oriented Agent Simulations: From Theory to Implementation", in: M. Ghallab, C.D. Spyropoulos, N. Fakotakis, M. Avouris (eds.), 18th European Conference on Artificial Intelligence (ECAI'08), Frontiers in Artificial Intelligence Applications, IOS Press, p. 383-387, ISBN 978-1-58603-891-5.

P. MATHIEU and S. PICAULT (2005), "Towards an interaction-based design of behaviors", in: M.-P. Gleizes (ed.), Proceedings of the The Third European Workshop on Multi-Agent Systems (EUMAS'2005).

U. WILENSKY (1999), NetLogo. https://ccl.northwestern.edu/netlogo Center for Connected Learning and Computer-Based Modeling, Northwestern University. Evanston, IL.

How to rebuild

If you introduce changes in the Java files, the jar file can be rebuilt using Maven:

mvn clean package
mv target/ioda.jar .
mvn clean

License

Centre de Recherche en Informatique, Signal et Automatique de Lille (CRIStAL), UMR CNRS 9189

Université de Lille – Cité Scientifique, F-59655 Villeneuve d'Ascq Cedex, FRANCE.

The IODA NetLogo extension is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

The IODA NetLogo extension is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with IODA NetLogo extension.
If not, see https://www.gnu.org/licenses.

cristal-smac/ioda