File list--for each parameter tested, there is one:

• Raw data file containing network information at each timestep
• Raw data file containing health information for agents at each timestep
• Full model with graphs

Model Information

The default parameters for this model are as follows:

• Population = 1000
• Timesteps = 100
• CFR = 0.65
• Initial number of infected = 10
• Health workers in population = 10

Assumptions for model:

• Population is ignorant of ebolavirus
• Deceased individalas are buried the same day of death
• Deceased are more infectious than infected individuals
• Healers / Doctors will always seek to help infected

Tested parameters:

• Transmission from dead --> susceptible (CORPSE)
• Transmission from infected --> susceptible (INFECTED)
• Case Fatality Rate effect on transmission (CFR)
• Amount of health workers effect on transmission (HEALTHWORKER)

Model Description

• Model is here run to 100 timesteps (It is recommended to run to completion, and this can be done with a while loop. However this may cause model to take > 5 hours to run each tested parameter)
• 100 iterations are done for each tested parameter

1. Doctor Movements

• Doctor / Healthcare worker checks for connection to infected
• If doctor is attached to an infected, no movement occurs
• If doctor is NOT attached to an infected, doctor will find an infected person in the population to attach to
• Doctor will attempt to heal infected individaul

2. Latent Individuals Checked

• Exposed individuals are identified, checked and updated in the exposed registry
• Individuals exposed for 8 days become symptomatic and enter infected class

3. Infected Individuals Checked

• Infected individauls are identified, checked and updated in the infected registry
• At 3 days in, infected indivudals can start to feel better and enter recovered class
• Past 3 days in, infected individuals can start dying
• At 8 days in, infected individuals die

4. New Infections

• If infected person connected to a susceptible person, infection can occur at previously set probability
• If dead person connected to a susceptibe person, infection can occur at previously set probability

5. Burials

• Dead individauls are removed from the network

Results

Example of one simulation of a community at time step 1

Example of case data obtained from 100 simulations of control parameters

The blue line is the average number of cases present at each time step

Example of death data obtained from 100 simulations of control parameters

The blue line is the average number of deaths at each time step.

Testing a range of Case Fatality Rates

Normally distributed, W = 0.92105, p-value = 0.4385
AOV: F = 3.904, p = 0.0956, No statistical difference

Normally distributed : W = 0.97896, p-value = 0.9576
AOV: F = 0.125, p = 0.736, No statistical difference

This is result is interesting. One could expect a higher case fatality rate resulting in a higher amount of deaths. However, this lack of difference could be due to higher CFR rates being detrimental to viral success. A higher death rate could decrease the chances the virus has to infect another individual.

Changing amount of treatment available in the community

Normally Distributed: W = 0.93463, p-value = 0.4949
AOV: F = 0.469, p = 0.513, No statistical difference

Normally distributed :W = 0.97536, p-value = 0.9356
AOV: F = 0.358 p = 0.566, No statistical difference

Testing a range of infection rates from corpses

Shapiro-wilks: W = 0.93404, p-value = 0.4888
AOV: F = 380.9, p < 0.001, means of case numbers statistically different

Shapiro-Wilks: W = 0.93291, p-value = 0.4771
AOV: F = 349.1, p < 0.001, means of deaths statistically different

Testing a range of infection rates from live hosts

Shapiro-wilks: W = 0.9631, p-value = 0.8206
AOV: F = 102.8, p <.001, means of case numbers statistically different

Shapiro-Wilks: W = 0.95509, p-value = 0.7287
AOV: F = 95.55, p < 0.001, means of deahts statistically different