This is a JavaScript implementation of the classical infectious disease model SEIR (Susceptible → Exposed → Infected → Removed) used in epidemiology to simplify the modelling of infectious diseases. Learn more
Most of this code was copied from the wonderfull Epidemic Calculator and rewritten in TypeScript and published on npm.
Can be used in node.js and in the browser.
Zero dependencies.
npm install seir
import SeirModel from 'seir';
const covid19 = new SeirModel({
r0: 2.2,
dDeath: 32 - 2.9,
dIncubation: 5.2,
dInfectious: 2.9,
dRecoveryMild: 14 - 2.9,
dRecoveryServere: 31.5 - 2.9,
dHospitalLag: 5,
cfr: 0.02,
pServere: 0.2,
dt: 2,
});
const timeline = covid19.calculate({
population,
r0ReductionPercent: 10,
r0ReductionDay: 5,
days: 300,
});
/*
[
{ deaths: 0, hospitalized: 0, recovered: 0, totalInfected: 1, infected: 1, exposed: 2 },
{ deaths: 0, hospitalized: 0, recovered: 1, totalInfected: 4, infected: 3, exposed: 5 },
...
]
*/
console.log(timeline);| Name | Required | Type | Default | Description |
|---|---|---|---|---|
| r0 | true | Number | - | Maximnum value of R0 if no measures are taken |
| dDeath | true | Number | - | Time from start of infection to death |
| dIncubation | true | Number | - | Length of incubation period |
| dInfectious | true | Number | - | Duration patient is infectious |
| dRecoveryMild | true | Number | - | Recovery time for mild cases |
| dRecoveryServere | true | Number | - | Recovery time for servere cases |
| dHospitalLag | true | Number | - | Time to hospitalization |
| cfr | true | Number | - | Case fatality rate |
| pServere | true | Number | - | Hospitalization rate |
| duration | false | Number | 7 * 12 * 1e10 | Duration of simulation |
| dt | false | Number | 1 | Time step size (days) |
| integrator | false | String | RK4 | Integrator to use. Available are Euler, Midpoint, Heun, K3, SSP33, SSP43, RK4, RK38 |
| Name | Required | Type | Default | Description |
|---|---|---|---|---|
| population | true | Number | - | Population of the location to run the simulation on |
| initiallyInfected | false | Number | 1 | Number of initially infected people |
| initiallyExposed | false | Number | 0 | Number of initially exposed people |
| r0ReductionPercent | false | Number | 0 | How much the r0 value should be reduced (for example because of measures) |
| r0ReductionDay | false | Number | 0 | On which the the reduction should take place |
| days | false | Number | 110 | For how many days the calculation should return a value |
Gabriel Goh for creating the Epidemic Calculator and therefore implementing pretty much all of these calculations. He also acknowledged Steven De Keninck for RK4 Integrator. Chris Olah, Shan Carter and Ludwig Schubert wonderful feedback. Nikita Jerschov for improving clarity of text. Charie Huang for context and discussion.
MIT