Skip to content

Kaplan-Meier-Estimator also known as the product limit estimator.

License

Notifications You must be signed in to change notification settings

fllstx/kaplan-meier-estimator

Repository files navigation

kaplan-meier-estimator

A JavaScript implementation of the Kaplan-Meier-Estimator also known as the product limit estimator.

Installation

node

npm install --save @fullstax/kaplan-meier-estimator

Browser

<script src="https://unpkg.com/kaplan-meier-estimator@latest/lib/kaplan-meier-estimator.umd.js">

Usage

esm/typescript usage

See examples/node for an example implementation.

import { compute } from '@fullstax/kaplan-meier-estimator';

const timeToEvents = [1, 12, 22, 29, 31, ];
const events = [false, true, false, true, true, ];

const kmData = compute(timeToEvents, events);
console.table(kmData);

CommonJs is also supported:

const { compute } = require('@fullstax/kaplan-meier-estimator');

browser usage

See examples/browser for an example implementation. To server this example jsu start d webserver in the root of this repository.

<script src="https://unpkg.com/@fullstax/kaplan-meier-estimator/lib/kaplan-meier-estimator.umd.js">
const timeToEvents = [1, 12, 22, 29, 31, ];
const events = [false, true, false, true, true, ];

const kmData = KME.compute((timeToEvents, events);

console.table(kmData);

example output

┌─────────┬───────┬──────┬───────┐
│ (index) │  rate │ time │ event │
├─────────┼───────┼──────┼───────┤
│    0    │   1   │  1   │ false │
│    1    │ 0.929 │  12  │ true  │
│    2    │ 0.929 │  22  │ false │
│    3    │ 0.851 │  29  │ true  │
│    4    │ 0.774 │  31  │ true  │
│    …    │   …   │   …  │   …   │
└─────────┴───────┴──────┴───────┘

Example

Data

# Time to Event event > S(t)
1 1 false 1
2 12 true 0.93
3 22 false 0.93
4 29 true 0.85
5 31 true 0.77
6 36 false 0.77
7 38 false 0.77
8 50 false 0.77
9 60 false 0.77
10 61 true 0.64
11 70 true 0.52
12 88 false 0.52
13 99 false 0.52
14 110 false 0.52
15 140 false 0.52

Plot

Kaplan-Meier example plot, CC-0

License

Licensed 2022 by fullstax GmbH & Co. KG under a MIT License.