sliding-window-max

Given a stream of data, this algorithm returns (for every added value) the current max value.

It uses a strategy that:

• Stores at most a number of values defined by the window size
• Performs a rolling max calculation

Example

const SlidingWindowMax = require('sliding-window-max')
 
const windowSize = 3
 
const slidingWindowMax = new SlidingWindowMax(windowSize)
slidingWindowMax.add(0) // returns null
slidingWindowMax.add(5) // returns null
slidingWindowMax.add(10)// returns 10
slidingWindowMax.add(7) // returns 10
slidingWindowMax.add(8) // returns 10
slidingWindowMax.add(5) // returns 8
slidingWindowMax.add(3) // returns 8
slidingWindowMax.add(4) // returns 5
//etc ...

Reference

new SlidingWindowMax(windowSize, options)

add(value)

Evaluate a new value and returns the calculated max value

Credits

I made this algorithm starting from this code https://github.com/chihungyu1116/leetcode-javascript/blob/master/239 Sliding Window Maximum.js. My requirements were a bit different. The leetcode algorithm requires that all the data are known before it starts. With sliding-window-max you can:

• evaluate the data as soon as they are produced
• evaluate big array using less memory

Changelog

• 0.x - Beta version
• 1.0 - First official version
• 1.1 - Migrates to gh-workflows; Upgrades deps

## Contributors

• chrvadala (author)