Implementing i-state configuration models for population dynamics: an object-oriented programming approach

Demographic models describe population dynamics in terms of the distribution of individuals among categories (e.g., age or size classes); such models are called i-state distribution models. In contrast, i-state configuration models describe population dynamics by simulating the birth, development, and eventual death of each individual in the population. Such models (also referred to, less precisely, as individual-based models) are necessary when interactions among individuals make the assumptions of i-state distribution models untenable. The basic components of an i-state configuration model are a set of individuals (each characterized by its i-state), an interaction structure, and an environment. Each of these components changes dynamically as a function of the others. The implementation of i-state distribution models is familiar; here we present a general framework, based on object-oriented programming (OOP), for the numerical implementation of i-state configuration models. The individuals, interaction structure, and environment are all defined as objects. A special object called the simulator transfers information among these objects as needed. The properties of OOP (data protection, inheritance, polymophism, modularity) lend themselves naturally to i-state configuration simulations.