Chapter 5
Agents in Social Science
5.1
Sugarscape Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
92
5.1.1
Evolution from Bottom-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
5.1.2
Distribution of Wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
94
5.1.3
Location Is Important! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
95
5.1.4
Find Agents around Me . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.1.5
Handle Multiple ‘Eaten’ Requests . . . . . . . . . . . . . . . . . . . . . . . 105
5.1.6
Change Starting Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.2
Modeling Social Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.2.1
Set Up a Recurring Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
5.2.2
Assigning Conditions with Functions . . . . . . . . . . . . . . . . . . . 113
5.2.3
Using Dynamic Arrays and Data Structures . . . . . . . . . . . . 113
5.2.4
Creating Local Dynamic Arrays . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3
Modeling Pedestrians in Crowds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3.1
Calculate Movement toward Other Agents . . . . . . . . . . . . . . 116
5.3.2
Entering and Exiting Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Decentralized control is an important aspect of self-organizing systems. Behavior of insect colonies is studied to deduce how, despite working independently,
a colony can work so efficiently. Termites and ants are examples of this. In ant
colonies, Wilson and H¨ olldobler [202] argued that every ant follows a particular ‘rule of thumb’ while making decisions based on local stimuli. When all
actions are put together, this emerges into complicated but precise execution
plans for the colony. This behavior has evolved over millions of years, driven
by natural selection.
Growing artificial agent societies is a useful technique to study how societies are created and thrive in changing real world conditions. Computer
simulations can be used to study organisms interacting together in a safe
environment, validated with experimental data.
Social scientists have used agent-based models in various political, ecological and economic scenarios. Here, agent-based models are ideal for understanding models involving individuals who interact and produce emergent
phenomena. Writing agent-based models begins with assumptions on the interactions among agents. The agents are then simulated, producing and modifying variables depending on these interactions and time. Simulations are used
as an addition to scientific analysis from deductions and inductions. Here sim87
Agents in Social Science
5.1
Sugarscape Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
92
5.1.1
Evolution from Bottom-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
93
5.1.2
Distribution of Wealth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
94
5.1.3
Location Is Important! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
95
5.1.4
Find Agents around Me . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
5.1.5
Handle Multiple ‘Eaten’ Requests . . . . . . . . . . . . . . . . . . . . . . . 105
5.1.6
Change Starting Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.2
Modeling Social Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
5.2.1
Set Up a Recurring Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112
5.2.2
Assigning Conditions with Functions . . . . . . . . . . . . . . . . . . . 113
5.2.3
Using Dynamic Arrays and Data Structures . . . . . . . . . . . . 113
5.2.4
Creating Local Dynamic Arrays . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3
Modeling Pedestrians in Crowds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114
5.3.1
Calculate Movement toward Other Agents . . . . . . . . . . . . . . 116
5.3.2
Entering and Exiting Agents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Decentralized control is an important aspect of self-organizing systems. Behavior of insect colonies is studied to deduce how, despite working independently,
a colony can work so efficiently. Termites and ants are examples of this. In ant
colonies, Wilson and H¨ olldobler [202] argued that every ant follows a particular ‘rule of thumb’ while making decisions based on local stimuli. When all
actions are put together, this emerges into complicated but precise execution
plans for the colony. This behavior has evolved over millions of years, driven
by natural selection.
Growing artificial agent societies is a useful technique to study how societies are created and thrive in changing real world conditions. Computer
simulations can be used to study organisms interacting together in a safe
environment, validated with experimental data.
Social scientists have used agent-based models in various political, ecological and economic scenarios. Here, agent-based models are ideal for understanding models involving individuals who interact and produce emergent
phenomena. Writing agent-based models begins with assumptions on the interactions among agents. The agents are then simulated, producing and modifying variables depending on these interactions and time. Simulations are used
as an addition to scientific analysis from deductions and inductions. Here sim87