A binary heap, suppose heap has n elements

isEmpty - O(1)
decreaseKey - O(log(n))
increaseKey - O(log(n))
insert - O(log(n))
delete - O(log(n))
from - O(n)
extract - O(log(n))
peek - O(1)

The heap is implemented internally as a binary tree represented by an array, and a map we maintain allows us O(1) lookup of elements. As such it has a greater memory overhead, and slightly lengthier insert, extract, etc. with the trade off of O(log(n)) delete, increase and decrease operations.

You can use it like so

const PriorityQueue = require("binaryheaps");

const arr = [2, 3, 1, 5, 6];
const cmp = (a, b) => a >= b ? 1 : -1;

const p = PriorityQueue.from(arr, cmp);

const largest = p.extract();

console.log(largest) // 6

In general you can either call new PriorityQueue(cmp) where cmp is a function (a: T, b: T) => number which returns 1 if a > b, 0 if a === b, else -1, where T is the type of the elements inserted into your queue.

Alternatively you can use the static from method as above which takes as its first argument an iterable.