https://github.com/tianyishi2001/kdtree

k-dimensional tree in Rust
https://github.com/tianyishi2001/kdtree

Last synced: about 1 month ago
JSON representation

k-dimensional tree in Rust

• Host: GitHub
• URL: https://github.com/tianyishi2001/kdtree
• Owner: TianyiShi2001
• License: mit
• Created: 2020-12-23T09:27:26.000Z (almost 4 years ago)
• Default Branch: main
• Last Pushed: 2020-12-24T01:59:49.000Z (almost 4 years ago)
• Last Synced: 2024-09-15T02:56:34.462Z (2 months ago)
• Language: Rust
• Size: 12.7 KB
• Stars: 4
• Watchers: 3
• Forks: 1
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# kdtree

k-dimensional tree data structure implemented with const generics, used for finding k-nearest neighbours (KNN).

## Example Usage

```rust

use kdt::*;

use ordered_float::OrderedFloat;

use rand::{thread_rng, Rng};

fn main() {

    let mut points = {

        let mut rng = thread_rng();

        (0..100)

            .map(|_| {

                Point([

                    rng.gen_range(-50.0..50.0),

                    rng.gen_range(-50.0..50.0),

                    rng.gen_range(-50.0..50.0),

                ])

            })

            .collect::>()

    };

    let kdt = KdTree::from_slice(&mut points);

    let query = Point([0.0, 0.0, 0.0]);

    let nearest = kdt

        .k_nearest_neighbors(&query, 10)

        .into_iter()

        // each point is returned as a reference. In most use cases you don't need to `clone`

        .map(|(dist, point)| (dist, point.clone()))

        // by default results are sorted in descending order of squared Eucledian distance to the query point

        .rev()

        .collect::>();

    // compute by brutal force

    let mut expected = points

        .into_iter()

        .map(|p| (p.squared_eucledian(&query), p))

        .collect::>();

    expected.sort_unstable_by_key(|p| OrderedFloat(p.0));

    assert_eq!(&nearest[..], &expected[..10]);

}

```