Order this book
University of Salzburg
It is, therefore, anything but easy to find adequate niches where to place a book with a similar intention. The book Agent-based Spatial Simulation with NetLogo, Volume 1: Introduction and Bases, edited by Arnaud Banos, Christophe Lang and Nicolas Marilleau is such an attempt (Volume 2 will follow in one of the next issues).
Volume 1 is organised in six chapters. Four of them are dealing with the theory and methodology of agent-based modelling, and the last two chapters offer insights into a complementary method and a particular utilisation. Chapter 1 introduces the agent approach by focusing on the agent paradigm and trends in spatial modelling. Chapter 2 presents an overview of how to formalise agent-based models by delineating the ODD (Overview, Description, Details) style of documentation, Unified Modelling Language (UML) and Agent Modelling Language (AML). An introduction to the NetLogo software platform is given in Chapter 3, followed by a presentation of visualisation tools which help explore simulation results (Chapter 4). Chapter 5 provides the reader with the system dynamics approach that can also be utilised within the NetLogo simulation platform. System dynamics with its analysis of (aggregated) stocks and their relationships can be understood as a complimentary technique to the ABM approach. The last chapter of Volume 1 describes different possibilities to create a collaborative environment either for model developers or model applicants. NetLogo’s server-client HubNet tool and the open source PAMS portal are introduced with concrete examples of how it can be used to enhance participation and cooperation among and between researchers and practitioners.
Two general aspects characterise the book: first, it is not dedicated to beginners in the field of modelling and simulation, neither with respect to ABM theory and methodology nor with respect to the use of NetLogo. Readers should have to be familiar, at least to some extent, with the modelling and simulation paradigm in scientific analysis and also should have some experience with programming languages in general and NetLogo programming in particular. This fact, however, does not discredit the book.
Secondly, the book struggles with finding a clear and distinct identity. It wants to be an introduction to agent-based modelling, and agent-based spatial modelling, and agent-based spatial modelling with NetLogo. It is not to say that it failed completely to achieve this aspiration, but one can feel this struggle throughout the book. One example is Chapter 1.3 Agents and the major trends in spatial modelling. The authors tried to delineate major trends in spatial modelling at only one page and got lost with a definition of the term “model”. This rough overview is followed by a slightly more comprehensive review of four “instantiations of models” of which only one (models of spatial interaction) actually refers to spatial modelling. The other three instantiations – statistical and econometric models, optimization models, and simulation models – have no explicit reference to space in any specific way.
A lot of particularities, hints, and recommendations of how to create and develop models with NetLogo are given, and they are helpful in solving problems of model verification in particular and model progress in general. Visual explorations or an ABM representation of a system dynamics panic model are examples of the strength of the volume. What is missing, however, is a similarly extensive description and discussion of modelling social phenomena in a spatially explicit way. The topic of modelling and simulating geographical and social space is definitely underrepresented which is surprising when taking the title of the book into account.
A very interesting chapter is Chapter 6 which describes possibilities to involve stakeholders of different domains – those who are interested in understanding the procedures and results of simulation models, and those who are generating simulation models correctly and efficiently in terms of verification, calibration, and empirical validation – in one collaborative modelling environment. Another stimulating chapter is Chapter 5 which provides the reader with a good overview of the idea of system dynamics and their procedures implemented in NetLogo. This is followed by a comparison with an agent-based approach that deals with the same problem differently. Unfortunately, the chapter abruptly finishes with an unspecified suggestion that the reader now should develop an agent-based model of her/his own interest in the system dynamics realm within the NetLogo environment. This suggestion can be executed only by an experienced NetLogo modeller; the material provided in the book does not suffice to complement this task.
To conclude: readers who want to know basics or particularities about agent-based spatial modelling and simulation will be disappointed, but readers with some level of experience in NetLogo programming and interested in collaborative modelling environments and coupling system dynamics with ABM will find useful material.
HEPPENSTALL A., Crooks A. T., See L. M. and M. Batty (eds.) (2012). Agent-Based Models of Geographical Systems. Springer: Dordrecht Heidelberg London New York.
O’SULLIVAN D. and G. L. W. Perry (2013). Spatial Simulation. Exploring Pattern and Process. Wiley-Blackwell: Chichester.
RAILSBACK S. F. and V. Grimm (2012). Agent-Based and Individual-Based Modeling. Princeton University Press: Princeton and Oxford.
SIEGFRIED R. (2014). Modeling and Simulation of Complex Systems. Springer: Wiesbaden.
WILENSKY U. and W. Rand (2015). An Introduction to Agent-Based Modeling. The MIT Press: Cambridge, London.
Return to Contents of this issue
© Copyright JASSS, 2018