ABMSync: An Object-Oriented Reimplementation of PS-I
R. Fradin,Steve Pelhan,B. Silverman
Abstract:The purpose of this project is to create a viable option for an industry standard for Agent Based Modeling (ABM) applications. By carefully planning our approach with an emphasis on an intuitive object model, readability, and flexibility, we intend to incorporate ease of use and functionality into one Python application. The current standards have generally only been able to incorporate one of these two important aspects. However, one open-source program, Political Science-Identity (PS-I) attempts to do both. As such, we will be using this program as our guide in terms of functionality and top level usability. However, the program is not without major designs flaws. Our Senior Design project will attempt to eradicate these flaws by developing a solid base from which PS-I can be reimplemented and improved. We will do this by first implementing a base on which most ABM applications can be created and then we will recreate PS-I as a module on top of that program. Related Work Agent Based Modeling is a computational model that runs a simulation to produce some event driven output. According to Borshchev and Filippov, "Modeling is a way of solving problems that occur in the real world. It is applied when prototyping or experimenting with the real system is expensive or impossible." Also, "a simulation model may be considered as a set of rules (e.g. equations, flowcharts, state machines, cellular automata) that define how the system being modeled will change in the future, given its present state." It is also very important to note that "simulation is the process of model 'execution' that takes the model through (discrete or continuous) state changes over time" (Borshchev & Filippov, 2004). Agent Based Modeling (ABM) differs from other well known methods of modeling, namely, System Dynamics and Discrete Event Modeling. Whereas those two types of modeling have a large, centralized "hub" that determines how the model evolves, ABM introduces a new concept of individual agents, that each have rules for how they will evolve. However, the idea of what exactly constitutes an agent is up to debate. Borshchev and Filippov say, "there are no universally accepted definitions in this area, and people still discuss what kind of properties an object should have to 'deserve' to be called an 'agent': proand re-activeness, spatial awareness, ability to learn, social ability, 'intellect,' etc...but we would like to stress just one feature of agent based models: they are essentially decentralized. (original emphasis)" (Borshchev & Filippov, 2004) This decentralization is what gives ABM its flexibility and power. "Agent based modeling is for those who wish to go beyond the limits of [System Dynamics] and [Discrete Event] approaches especially in the case the system being modeled contains active objects (people, business units, animals, vehicles, or projects, stocks, products, etc.) with timing, event ordering or other kind of individual behavior." (Borshchev & Filippov, 2004) What is an Agent: Now we know that "Agent-based models are (computational) models of a heterogeneous population of agents and their interactions. The result of the micro-level interactions can be interesting macro-level behavior like cooperation, segregation, fashion, culture, etc" (Agent Based Modeling FAQ, 2008), but it may still be unclear what an agent is. Macal and North describe an agent as "a discrete entity with its own goals and behavior" and it must be "autonomous, with a capability to adapt and modify its behavior" (Macal & North, 2006). For the purpose of the programs we have studied, an agent can represent many things including, "individuals, households, firms, and nations," and can be encoded to act in many different ways, "such as changing the shared environment by harvesting or pollution, exchange of information and resources, and by imitation." Perhaps the most important feature of ABM is the rules used to define how the agents interact. However, defining what those decision rules are can be a very difficult question. The Open ABM Consortium answers it this way, "instead of assuming selfish individuals with perfect information, we often assume agents who have simple heuristics of satisfying behavior."