sklears-manifold

Latest release: 0.1.0-beta.1 (January 1, 2026). See the workspace release notes for highlights and upgrade guidance.

Overview

sklears-manifold implements manifold learning, nonlinear dimensionality reduction, and embedding algorithms mirroring scikit-learn’s manifold module.

Key Features

• Algorithms: t-SNE, UMAP-compatible neighbors, Isomap, Locally Linear Embedding, Spectral Embedding, MDS.
• Performance: Barnes-Hut and FFT-based t-SNE, GPU nearest neighbors, and multithreaded eigen solvers.
• Visualization: Embedding utilities that integrate with sklears-inspection and Python plotting stacks.
• Pipeline Support: Works seamlessly with preprocessing, decomposition, and clustering crates.

Quick Start

use sklears_manifold::TSNE;
use scirs2_core::ndarray::Array2;

let x: Array2<f32> = // load dataset
    Array2::zeros((2000, 128));

let tsne = TSNE::builder()
    .n_components(2)
    .perplexity(30.0)
    .learning_rate(200.0)
    .n_iter(1000)
    .build();

let embedding = tsne.fit_transform(&x)?;

Status

• Validated by the workspace’s 11,292 passing tests for 0.1.0-beta.1.
• Performance parity (and in many cases superiority) compared with scikit-learn’s manifold implementations.
• Upcoming tasks (GPU UMAP, streaming embeddings) tracked in TODO.md.