Chalarangelo/30-seconds-of-code

View on GitHub
content/snippets/js/s/heapsort.md

Summary

Maintainability
Test Coverage
---
title: Heap sort
type: story
language: javascript
tags: [algorithm,array,recursion]
cover: building-blocks
excerpt: Sort an array of numbers, using the heapsort algorithm.
listed: true
dateModified: 2023-12-16
---

## Definition

[Heapsort](https://en.wikipedia.org/wiki/Heapsort) is a **comparison-based sorting algorithm**. Heapsort can be thought of as an improved [selection sort](/js/s/selection-sort). The improvement consists of the use of a **heap data structure** instead of a linear-time search to find the maximum or minimum element.

## Implementation

- Use recursion.
- Use the spread operator (`...`) to clone the original array, `arr`.
- Use closures to declare a variable, `l`, and a function `heapify`.
- Use a `for` loop and `Math.floor()` in combination with `heapify` to create a max heap from the array.
- Use a `for` loop to repeatedly narrow down the considered range, using `heapify` and swapping values as necessary in order to sort the cloned array.

```js
const heapsort = arr => {
  const a = [...arr];
  let l = a.length;

  const heapify = (a, i) => {
    const left = 2 * i + 1;
    const right = 2 * i + 2;
    let max = i;
    if (left < l && a[left] > a[max]) max = left;
    if (right < l && a[right] > a[max]) max = right;
    if (max !== i) {
      [a[max], a[i]] = [a[i], a[max]];
      heapify(a, max);
    }
  };

  for (let i = Math.floor(l / 2); i >= 0; i -= 1) heapify(a, i);
  for (i = a.length - 1; i > 0; i--) {
    [a[0], a[i]] = [a[i], a[0]];
    l--;
    heapify(a, 0);
  }
  return a;
};

heapsort([6, 3, 4, 1]); // [1, 3, 4, 6]
```

## Complexity

The algorithm has an [average time complexity](/js/s/big-o-cheatsheet#array-sorting-algorithms) of `O(n log n)`, where `n` is the size of the input array.