#clustering #cluster #means

nightly kmeans

Small and fast library for k-means clustering calculations

2 releases

0.2.1 Jan 3, 2024
0.2.0 Oct 8, 2020
0.1.0 Jul 27, 2019

#893 in Algorithms

Download history 12/week @ 2023-12-11 17/week @ 2023-12-18 2/week @ 2023-12-25 30/week @ 2024-01-01 18/week @ 2024-01-08 7/week @ 2024-01-15 6/week @ 2024-01-22 18/week @ 2024-02-12 38/week @ 2024-02-19 67/week @ 2024-02-26 29/week @ 2024-03-04

152 downloads per month
Used in 2 crates

Apache-2.0

89KB
1K SLoC

kmeans

Current Crates.io Version docs

kmeans is a small and fast library for k-means clustering calculations. Here is a small example, using kmean++ as initialization method and lloyd as k-means variant:

use kmeans::*;

fn main() {
    let (sample_cnt, sample_dims, k, max_iter) = (20000, 200, 4, 100);

    // Generate some random data
    let mut samples = vec![0.0f64;sample_cnt * sample_dims];
    samples.iter_mut().for_each(|v| *v = rand::random());

    // Calculate kmeans, using kmean++ as initialization-method
    let kmean = KMeans::new(samples, sample_cnt, sample_dims);
    let result = kmean.kmeans_lloyd(k, max_iter, KMeans::init_kmeanplusplus, &KMeansConfig::default());

    println!("Centroids: {:?}", result.centroids);
    println!("Cluster-Assignments: {:?}", result.assignments);
    println!("Error: {}", result.distsum);
}

Datastructures

For performance-reasons, all calculations are done on bare vectors, using hand-written SIMD intrinsics from the packed_simd crate. All vectors are stored row-major, so each sample is stored in a consecutive block of memory.

Supported variants / algorithms

  • lloyd (standard kmeans)
  • minibatch

Supported centroid initialization methods

  • KMean++
  • random partition
  • random sample

Dependencies

~3.5MB
~67K SLoC