kdbush

What is kdbush?

kdbush is a high-performance JavaScript library for creating and querying static spatial indexes using a k-d tree. It is particularly useful for handling large sets of 2D points and performing fast nearest neighbor searches.

What are kdbush's main functionalities?

Creating a spatial index

This feature allows you to create a spatial index from an array of 2D points. The index can then be used for efficient spatial queries.

const KDBush = require('kdbush');
const points = [
  [20, 30],
  [40, 50],
  [10, 15],
  [25, 35]
];
const index = new KDBush(points);

Performing a range search

This feature allows you to perform a range search on the spatial index. The `range` method returns all points within the specified bounding box.

const KDBush = require('kdbush');
const points = [
  [20, 30],
  [40, 50],
  [10, 15],
  [25, 35]
];
const index = new KDBush(points);
const results = index.range(10, 10, 30, 40);

Performing a nearest neighbor search

This feature allows you to perform a nearest neighbor search on the spatial index. The `within` method returns all points within a specified radius from a given point.

const KDBush = require('kdbush');
const points = [
  [20, 30],
  [40, 50],
  [10, 15],
  [25, 35]
];
const index = new KDBush(points);
const nearest = index.within(20, 30, 10);

Other packages similar to kdbush

rbush

rbush is a high-performance JavaScript library for 2D spatial indexing of points and rectangles using an R-tree. It is similar to kdbush but uses an R-tree data structure, which can be more efficient for certain types of spatial queries, especially when dealing with rectangles.

geokdbush

geokdbush is a geospatial extension of kdbush that adds support for geographic coordinates (latitude and longitude). It provides similar functionality to kdbush but is specifically designed for geospatial data, making it more suitable for applications involving geographic information.

flatbush

flatbush is a fast static spatial index for 2D points and rectangles, similar to kdbush. It uses a flat R-tree data structure and is optimized for read-heavy workloads. It can be more efficient than kdbush for certain types of spatial queries, particularly when dealing with large datasets.

README

KDBush

A very fast static spatial index for 2D points based on a flat KD-tree. Compared to RBush:

• Points only — no rectangles.
• Static — you can't add/remove items after initial indexing.
• Faster indexing and search, with lower memory footprint.
• Index is stored as a single array buffer (so you can transfer it between threads or store it as a compact file).

If you need a static index for rectangles, not only points, see Flatbush. When indexing points, KDBush has the advantage of taking ~2x less memory than Flatbush.

Usage

// initialize KDBush for 1000 items
const index = new KDBush(1000);

// fill it with 1000 points
for (const {x, y} of items) {
    index.add(x, y);
}

// perform the indexing
index.finish();

// make a bounding box query
const foundIds = index.range(minX, minY, maxX, maxY);

// map ids to original items
const foundItems = foundIds.map(i => items[i]);

// make a radius query
const neighborIds = index.within(x, y, 5);

// instantly transfer the index from a worker to the main thread
postMessage(index.data, [index.data]);

// reconstruct the index from a raw array buffer
const index = KDBush.from(e.data);

Install

Install with NPM: npm install kdbush, then import as a module:

import KDBush from 'kdbush';

Or use as a module directly in the browser with jsDelivr:

<script type="module">
    import KDBush from 'https://cdn.jsdelivr.net/npm/kdbush/+esm';
</script>

Alternatively, there's a browser bundle with a KDBush global variable:

<script src="https://cdn.jsdelivr.net/npm/kdbush"></script>

API

new KDBush(numItems[, nodeSize, ArrayType])

Creates an index that will hold a given number of points (numItems). Additionally accepts:

• nodeSize: Size of the KD-tree node, 64 by default. Higher means faster indexing but slower search, and vise versa.
• ArrayType: Array type to use for storing coordinate values. Float64Array by default, but if your coordinates are integer values, Int32Array makes the index faster and smaller.

index.add(x, y)

Adds a given point to the index. Returns a zero-based, incremental number that represents the newly added point.

index.range(minX, minY, maxX, maxY)

Finds all items within the given bounding box and returns an array of indices that refer to the order the items were added (the values returned by index.add(x, y)).

index.within(x, y, radius)

Finds all items within a given radius from the query point and returns an array of indices.

KDBush.from(data)

Recreates a KDBush index from raw ArrayBuffer data (that's exposed as index.data on a previously indexed KDBush instance). Very useful for transferring or sharing indices between threads or storing them in a file.

Properties

• data: array buffer that holds the index.
• numItems: number of stored items.
• nodeSize: number of items in a KD-tree node.
• ArrayType: array type used for internal coordinates storage.
• IndexArrayType: array type used for internal item indices storage.