Identifying at-risk individuals in social networks

To run a simulation type ./simulate-intervention.sh NAME. (It is an executable.) This creates a folder called NAME with subfolders corresponding to the maximum fraction of users being targeted. This fraction can be changed in simulate-itervention.sh.

Two subsubfolders are created in each subfolder, targeted and random. Targeted contains the results of the simulation targeted at-risk agents. Random contains the results of the simulation targeting a group of agents of the same size at the at-risk group chosen from the general population, including the at-risk subpopulation.

After the simulations _simulate-intervention.sh adds the directories where the data were stored to .gitignore.

1. A record of each simulation is saved as a tab-delimited text-file with four significant figures. For 200 time steps and 100,000 agents, this file is approximately 1 GB. Saving it in binary or as a 4 digit integer will save significant space.

2. This version uses forward Euler integration with a times step that is 1% of the time constant. This may be numerically unstable, increasingly so for longer simulations.

1. NumPy
2. SciPy
3. Matplotlib
4. Awesome Print
5. NetworkX
6. ProgressBar
7. Termcolor