Figure 1
Infection state progression, with arrows depicting the possible routes of progression through this scheme. Based on a SEIRD model, where all individuals are either susceptible (S) to the disease, exposed (E) to the disease, infected (I) by the disease, or who have recovered (R) or died (D) from the disease. Multiple compartments for infection account for cases of differing severity, and this network is highly configurable, and new compartments or connections can be added with ease.
Figure 2
Population structure in Epiabm. A population is formed of many microcells, which are grouped into cells (the largest spatial unit). Infected individuals may infect others within their cell through households and places, while infections are spread between cells according to a spatial kernel.
Figure 3
Spatial distribution of infected individuals within the population, at different time points during the simulation. Configured with a population of 10,000 people distributed across 200 cells, each containing 2 microcells with 5 households per microcell. One infected individual is initialised in the central cell, with mild infection status; the simulation is run for 80 days. Inter-cell infections only occur between nearby cells, allowing visualisation of the infection propagating through the simulation region over time.
Figure 4
A comparison of simulation outputs from pyEpiabm (a) and CovidSim (b), for an epidemic in Gibraltar initiated by 100 infected individuals. While the outputs are highly stochastic, a strong agreement is broadly observed.
Algorithm 1
Outline of the overall workflow for Epiabm.
Algorithm 2
Example of an infection sweep from Epiabm, handling infections occurring within a household.