Day0

Infectious0

StatusReady

Susceptible Vaccinated Exposed Infectious Recovered Dead Isolated Hospital bed No bed Large red dot = super-spreader

Population cumulative infected, no intervention

New cases / day no intervention

Effective R(t) R0·S/N, no intervention

Hospital load capacityno intervention

Epidemic Lab

1,500 people move between six towns. Pick a disease, switch interventions on and off, and compare against the dashed no-intervention curve. Click a person on the map to infect or isolate them.

Speed 2×

Random seed Initial cases

Click a person to

Disease

Basic reproduction number R0

Latent period Infectious period

Severity Travel between towns

Interventions

Vaccination

Coverage

Efficacy Waning half-life

Behaviour

Mask wearing Distancing

Lockdown: close travel between towns, stay home more

Auto-lockdown at infectious share

Quarantine & isolation

Isolate detected symptomatic cases

Detection rate Delay

Testing & contact tracing

Random tests, trace and quarantine contacts

Tests / 1000 / day Contacts traced

Hospitals

Beds Death risk w/o bed

Readouts

R0 (set / measured): —
Herd-immunity threshold: —
Peak infectious: —
Attack rate: —
Deaths: —
Immune now: —
Isolated now: —
Current R(t): —

What am I seeing?

Every dot is a person with a home, age group (child, adult, senior) and personal traits. About 6% are super-spreaders who transmit five times more; others spread less than average.
States run Susceptible → Exposed (infected, not yet contagious) → Infectious → Recovered or Dead. Some cases are asymptomatic: they spread half as much and are never detected by symptoms.
Contacts are found with a spatial hash grid. Each tick, a susceptible person within about 14 px of an infectious one has a chance of exposure that rises as they get closer and the longer they stay together; a flash line marks each transmission.
Towns are linked by roads. People make occasional trips, and a lockdown closes the roads and keeps people nearer home.
Herd immunity threshold = 1 − 1/R0. When more than that share is effectively immune, each case infects fewer than one other person and the outbreak dies out even though some people stay susceptible. A vaccine with efficacy e needs coverage of threshold / e.
R(t) is estimated from the daily new-case curve with a generation-interval kernel. Above 1 the outbreak grows, below 1 it shrinks. The measured R0 is inferred from how fast cases grew early on; clustering in households and towns usually makes it lower than the nominal R0 you set.
The dashed no-intervention baseline is a compartmental SEIRD differential-equation model with the same disease and a well-mixed population. Gaps between it and the dots show what your interventions achieved, plus effects of space and clustering.
Severe cases need a hospital bed. When demand exceeds capacity, those without a bed face a higher risk of dying. Flattening the curve keeps the load under the dashed capacity line.
Reproducible: the same seed and settings give the same run. Reducing vaccination coverage in mid-run removes protection from the latest-vaccinated people (a what-if, not a realistic event).
A teaching model with simplified biology; numbers are illustrative, not forecasts.