NetLogo User Community Models(back to the NetLogo User Community Models)
DrugPropagation
by Robert Rohrkemper and Josh Savory
(Submitted: 06/01/2004)
An Agentbased Model of Drug Propagation Using Three Populations by Bobby Rohrkemper and Josh Savory
WHAT IS IT?
The primary goal of the Drug Simulation is to model the propagation of drug addiction through an infected population. The secondary goal of this model is to determine whether making drugs illegal or legal is more effective in preventing the spread of drug addiction.
This model is based on a town with a population of 3,000 to 5,000 people. It observes the town over a short period of time, approximately 5 to 10 years, and looks at the change in the averageaddiction of the town’s population in this time period. Since the time frame of the model is so short the model assumes that the population size doesn't significantly change. This allows the model to focus purely on the effects of drugs and not have to simulate the effects of population growth. The town has the choice to make marijuana legal or illegal depending on which they feel will best control its spread. Making drugs illegal determines whether the town will combat the drug problem by arresting the users or by trying to rehabilitate them. They also have a sensitivity to the drug problem which determines how forcefully they will come down on the drug problem.
HOW IT WORKS
In this model we assume their are three different agents who affect the propagation of drug addiction. These three agents are: Susceptibles, TheLaw and AddictionVectors. Susceptibles are the segment of the population who are susceptible to drug addiction. TheLaw are agents who act to reduce the rate at which addiction spreads, i.e. cops, doctors, family members..... AddictionVectors are agents meant to simulate the factors that increases drug addiction i.e. peer pressure, physical addiction, the media... Each of these agents have rules that define what they should do when they are alone and when they interact other agents. The rules determining the agent’s interactions depend on whether marijuana is legal or illegal and the initial sensitivity of the population.
The Susceptibles start off with an initial addiction randomly disturbed over a power function. The purpose of the power function is to try and simulate a real world situation where you have a few Susceptibles with high addiction and many Susceptibles with a low addiction. The initial number of susceptibles is determined by the user. The Susceptibles addiction physical represents the probability that they will buy drugs when offered them. They start off at a random position and from their move about the screen randomly.
TheLaw like the Susceptibles starts off at random positions on the screen and move about randomly. Their population size is determined by the sensitivity of the overall population to the drug problem.
AddictionVectors are motionless agents and just wait until they have the chance to interact with the other agents. The number of AddictionVectors is determined by the user.
When Susceptibles and TheLaw encounter each other they will interact in one of three ways. If the addiction of the Susceptible is less then 1 minus the sensitivity of the population TheLaw will do nothing. If the addiction of the Susceptible is greater then 1 minus the sensitivity of the population and marijuana is illegal the Susceptible will be removed from the simulation. This is meant to simulate the arrest of the Susceptible. If the addiction of the Susceptible is greater then 1 minus the sensitivity of the population and marijuana is legal the Susceptible addiction will be decreased. This is meant to simulate the rehabilitation of the Susceptible.
When Susceptibles and AddictionVectors encounter each other the addiction of the Susceptible increases.
When TheLaw and AddictionVectors encounter each nothing happens. This could be another avenue to investigate in another simulation.
HOW TO USE IT
To use this model one must first set the Sensitivity of the population, the initialnumberofSusceptibles, the initialnumberofAddictionVectors and the Legality to the desired values. The Sensitivity describes how forcefully TheLaw will come down on the drug problem. The initialnumberofSusceptibles and the initialnumberofAddictionVectors determine the initial number of each subset of the population. The Legality determines whether marijuana is illegal or legal. After these values are set one then push setup to construct the requested scenario. When the user is read to start the simulation they must simply click on go.
THINGS TO NOTICE
Look for optimal values of averageaddiction. It is this quantity which we are trying to minimize.
THINGS TO TRY
Try legal and illegal scenarios, and explore how this effects the optimal values. Graph the averageaddiction as a function of the sensitivity as well as the initial number of susceptibles.
EXTENDING THE MODEL
There are many features which could be added to make the model more realistic. We have made a number of assumptions as outlined in our report.
NETLOGO FEATURES
We have not made use of any workarounds or special NetLogo code.
RELATED MODELS
This model is related to another project by Brad Boven, which can be found at max.cs.kzoo.edu/~bboven/drug_dynamics.pdf. However, his model was onedimensional and differentialequationbased, whereas ours is twodimensional and agentbased.
CREDITS AND REFERENCES
We would like to thank Gabor Csardi for his assistance with this project. Also, thanks to Dr. Peter Erdi for his instruction in complex systems theory.
