avila-clustering 0.1.0

State-of-the-art clustering algorithms for Rust - surpassing scikit-learn, HDBSCAN, and RAPIDS cuML
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44

//! ROCm (AMD) GPU acceleration for clustering

#[cfg(feature = "gpu-wgpu")]
use wgpu::{Device, Queue};

use crate::{ClusteringError, Result};

#[cfg(feature = "gpu-wgpu")]
pub struct RocmKMeans {
    n_clusters: usize,
    device: Device,
    queue: Queue,
}

#[cfg(feature = "gpu-wgpu")]
impl RocmKMeans {
    pub async fn new(n_clusters: usize) -> Result<Self> {
        // TODO: Initialize wgpu device
        unimplemented!("RocmKMeans::new")
    }

    pub fn fit(&self, data: &[Vec<f64>]) -> Result<RocmClusteringResult> {
        // TODO: Implement ROCm-accelerated KMeans using compute shaders
        unimplemented!("RocmKMeans::fit")
    }
}

#[cfg(feature = "gpu-wgpu")]
pub struct RocmClusteringResult {
    pub labels: Vec<usize>,
    pub centroids: Vec<Vec<f64>>,
}

#[cfg(not(feature = "gpu-wgpu"))]
pub struct RocmKMeans;

#[cfg(not(feature = "gpu-wgpu"))]
impl RocmKMeans {
    pub fn new(_n_clusters: usize) -> Result<Self> {
        Err(ClusteringError::GpuError(
            "GPU-wgpu support not enabled. Compile with --features gpu-wgpu".to_string(),
        ))
    }
}