Agent-based Spatial Simulation with NetLogo, Volume 2

Advanced Concepts

Arnaud Banos, Christophe Lang, Nicolas Marilleau

• Publisher: ISTE Press - Elsevier
• Year: 2016
• Print ISBN: 9781785481574
• eBook ISBN: 9780081010648
• Topic: Computer Science
• Subtopic: Programming
• Index: Computer Science

Whereas Volume 1 introduced the NetLogo platform as a means of prototyping simple models, this second volume focuses on the advanced use of NetLogo to connect both data and theories, making it ideal for the majority of scientific communities. The authors focus on agent-based modeling of spatialized phenomena with a methodological and practical orientation, demonstrating how advanced agent-based spatial simulation methods and technics can be implemented. This book provides theoretical and conceptual backgrounds, as well as algorithmic and technical insights, including code and applets, so that readers can test and re-use most of its content. - Illustrates advanced concepts and methods in agent-based spatial simulation - Features practical examples developed, and commented on, in a unique platform - Provides theoretical and conceptual backgrounds, as well as algorithmic and technical insights, including code and applets, so that readers can test and re-use most of its content

1

NetLogo, an Open Simulation Environment

Benoit Gaudou; Christophe Lang; Nicolas Marilleau; Guilhelm Savin; Sébastien Rey Coyrehourcq; Jean-Marc Nicod

Abstract

NetLogo is a generic simulation environment in the sense that it was not designed with any specific domain of application in mind. NetLogo offers a wide range of features and generic operators to its users. Additionally, to make up for any missing features, NetLogo is compatible with other platforms and libraries, as we will demonstrate throughout this book.

Keywords

Extensions; Java/Scala; Modeling Commons; NetLogo; OpenABM; Open Simulation Environment; Python; RungeKutta extension

1.1 Introduction to extensions in NetLogo

NetLogo is a generic simulation environment in the sense that it was not designed with any specific domain of application in mind. NetLogo offers a wide range of features and generic operators to its users. Additionally, to make up for any missing features, NetLogo is compatible with other platforms and libraries, as we will demonstrate throughout this book.

There is a vast library of extensions available to users, allowing them to integrate additional functionality that is not present in the native version of NetLogo, but which might nonetheless be necessary for the development of a given model. An official library of extensions is available on the official NetLogo Website. We will explore some of these extensions later in this chapter. But many modelers have also developed their own extensions to tackle specific problems that are of interest to them. These extensions are developed with an open Java API. We will discuss this in more detail in section 1.2.

Conversely, NetLogo can also be called and controlled by other programs, such as OpenMole¹, Python² and R³. To do this, NetLogo provides a Java API that allows models to be loaded, executed and gives access to their variables and methods. The usage of this API is presented in detail in section 1.3.

1.1.1 Examples of typical NetLogo extensions

There are many different types of extension. The GitHub page of the NetLogo platform⁴ gives one possible list of examples. This list distinguishes between internally developed extensions, which are included with the platform (e.g. GIS or network), and extensions developed by the community, which have to be installed manually (section 1.1.2). In Chapter 3 of Volume 1 [BAN 15a], we presented a number of these extensions (GIS and network) to showcase some of additional functionnalities of NetLogo.

Some of these extensions include language extensions, which allow the modeler to manipulate more complex object types than those natively present in NetLogo. Indeed, the language of the platform has relatively few complex structures (unlike most programming languages) and primitives for manipulating them. For instance, the array, table and matrix extensions are now included with NetLogo. However, extensions such as string and file are external.

More generally, the functionalities of the NetLogo language can be augmented with a wide range of extensions, for example to achieve better network management (network and nw), to provide more primitives for network analysis (additional metrics and indicators) or to integrate geographical data represented in vector form into NetLogo models (GIS). This is not an easy task, but this is absolutely necessary, as NetLogo models are natively based on a grid-based discrete environment. As another example, the SQL extension allows models to interact with a database by sending SQL-formatted queries and receiving data in response. Finally, the sound and MIDI extensions allow sounds to be integrated into NetLogo models.

There are a number of extensions that enable NetLogo to interact with other tools. This interaction can take various different forms: it might simply involve reading or writing files that are compatible with a third-party application. For example, it is possible to process image files (bitmap), tabular data (csv), Java system properties (Props), POV rays (RayTracing), VRML (VRML), NetCDF (NetCDF), etc. This allows modelers with different backgrounds to increase the realism of their simulations by exploiting real data in useful formats.

Other extensions allow deeper forms of interaction by directly integrating third-party functionality into NetLogo; for this kind of interaction, NetLogo must be able to connect with another application to send requests and retrieve results, such as Matlab (MATLAB), Prolog (NetProLogo), IODA [KUB 11] (IODA) and Graphstream⁵ (gs).

For example, the NetLogo language can be extended with primitives allowing it to benefit from the scientific calculation tool R [THI 10], and in particular to call R functions from within a NetLogo model. The R extension for NetLogo can be downloaded on the Netlogo-R-Extension Website⁶. It fulfills the task of communicating data between the two tools, and in particular performs type conversi...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Introduction
• 1: NetLogo, an Open Simulation Environment
• 2: Multiscale Modeling: Application to Traffic Flow
• 3: Macro Models, Micro Models and Network-based Coupling
• 4: Networking, Networks and Dynamic Graphs
• 5: Swarm Problem-Solving
• 6: Exploring Complex Models in NetLogo
• Conclusion
• Bibliography
• List of Authors
• Index