Designing Table-oriented Agent Simulation Framework

Abstract

In this paper, authors present table-oriented agent simulation framework and indicate internal design of the framework. The table-oriented agent simulation is a concept of all data including agents ' properties considered as cells of tables, and rules stored other spaces drive table-shaped data or agents. The framework for table-oriented agent simulation applies conventional thinking to create codes and tables, because of convenience to regard a name of a variable or a rule as the name of a file. The framework also provides a scaffold when a user executes the function of scaffolding with a template of a simulation process and initial data types. Finally, this paper shows an example on the framework to easily create a simulation program of Schelling's segregation model utilizing the template of Sugarscape model.