FAME-Io: Configuration tools for complex agent-based simulations

Abstract

We present FAME-Io, a Python package designed to help users and creators of agent-based simulation models (ABM) better manage the preparation and processing of their input and output datasets. The package was built with the needs of researchers in mind. FAME-Io was specifically developed to interface with the open framework FAME and is published under the open Apache-2.0 licence. The software offers various logging capabilities, shell-integrated help and documentation, as well as extensive pre-run integrity checks and helpful warning messages. It also allows individual data components to be easily extracted and used in secondary workflows. The code itself is operating system independent and follows best practices in software development. Test coverage, at the time of writing, is 92% and the project uses continuous integration and offers frequent releases. FAME-Io is designed as one of two main components of FAME, each addressing a particular aspect and user group within ABM development: - FAME-Io is a Python package for ABM users. It supports configuring complex simula- tions and managing of files associated with FAME. - FAME-Core is a Java library for ABM modellers. It supports developing and executing ABM simulations. These two modules interact with additional supporting components, like FAME-Mpi, FAME-Protobuf, FAME-Gui (in beta release) and FAME-Prepare. The demonstration project FAME-Demo provides a ready-made example for modellers to investigate and experiment with.