Agent-Based Modelling – Coursework 1 – 2023
Systematic Experimentation – SugarScape
Submission: Word or PDF document via Moodle
Late submission of this work will be subject to the standard penalties, according to UCL regulations.
task
Consider the Netlogo Models Library models “Sugarscape 1 Immediate Growback”, “Sugarscape 2 Constant Growback”, and “Sugarscape 3 Wealth Distribution”.
Choose at least two of these models, and use BehaviorSpace to perform systematic investigations of their behaviour as parameters are varied. This will require (a) choosing some aspect or aspects of the model behaviour to measure (e.g. final population, average vision, average metabolism, average wealth…) and potentially adding new reporters to the code (to-report … end) to calculate these metrics; and (b) choosing which parameters to alter and which values of those parameters to consider.
Present the results of your investigation in the form of a short report, with appropriate graphs and tables, and interpret these results to explain what has been learnt about the models’ behaviour, including some comparison of the behaviour of at least two versions of Sugarscape.
Weighting
This assessment makes up 50% of the marks for CASA0011.
Length
1500 words + graphs/tables. (Penalties for exceeding this limit will be in line with UCL regulations.)
Assessment
See the assessment criteria for Coursework 1, provided separately.
Structure
Your report should contain the following sections:
Aim – Explain what aspects of the models’ behaviour you intend to investigate.
Methods – Clearly and completely detail the experiments that you performed.
Results – Communicate the results of the experiments through graphs and tables.
Discussion – Interpret your results and compare the different models.
Conclusion – Summarise your main findings, relating back to your aims
Notes
-You are not expected to examine every possible value of every parameter. You should conduct experiments that can be comprehensively communicated within the 1500 words of this assignment.
-Any new reporters added to the code should be added to your report in an appendix. They will not be counted in the word limit.
-Graphs should never be copied and pasted from NetLogo. Instead the appropriate data should be exported to another piece of software (e.g. Excel) to create more attractive figures. This can be achieved by right clicking (or control-clicking, or similar) on a graph and selecting “Export…”
