FastMap

FastMap is an efficient algorithm that maps objects into a Euclidean space such that distances are preserved as much as possible, with a linear time complexity.

How It Works

FastMap uses a heuristic approach to approximate MDS linearly. It iteratively projects data points onto an orthogonal line defined by two distant pivot points, extracting coordinates one dimension at a time.

Why or When to Use

Use FastMap when you have a distance matrix and need a very fast (O(N)) linear dimensionality reduction, typically as a preprocessing step or for very large datasets where classical MDS is too slow.

Example

How-to (Code)

import * as druid from "@saehrimnir/druidjs";

const data = [
  /* ... multi-dimensional data ... */
];

// 1. Initialize the algorithm
const fastmap = new druid.FASTMAP(data);

// 2. Compute the projection
const projection = fastmap.transform();