Crate sklears_manifold 

Manifold learning algorithms (t-SNE, Isomap, etc.)

This module is part of sklears, providing scikit-learn compatible machine learning algorithms in Rust.

Re-exports

pub use tsne::TsneTrained;

pub use tsne::TSNE;

pub use isomap::Isomap;

pub use isomap::IsomapTrained;

pub use lle::LleTrained;

pub use lle::LocallyLinearEmbedding;

pub use mds::MdsTrained;

pub use mds::MDS;

pub use laplacian_eigenmaps::LaplacianEigenmaps;

pub use laplacian_eigenmaps::LaplacianTrained;

pub use umap::UmapTrained;

pub use umap::UMAP;

pub use diffusion_maps::DiffusionMaps;

pub use diffusion_maps::DiffusionMapsTrained;

pub use deepwalk::DeepWalk;

pub use deepwalk::DeepWalkTrained;

pub use dictionary_learning::DLTrained;

pub use dictionary_learning::DictionaryLearning;

pub use heavy_tailed_symmetric_sne::HeavyTailedSymmetricSNE;

pub use heavy_tailed_symmetric_sne::HeavyTailedSymmetricSneTrained;

pub use hessian_lle::HessianLLE;

pub use hessian_lle::HessianLleTrained;

pub use ltsa::LtsaTrained;

pub use ltsa::LTSA;

pub use minibatch_tsne::MBTSNETrained;

pub use minibatch_tsne::MiniBatchTSNE;

pub use minibatch_umap::MBUMAPTrained;

pub use minibatch_umap::MiniBatchUMAP;

pub use mvu::MvuTrained;

pub use mvu::MVU;

pub use node2vec::Node2Vec;

pub use node2vec::Node2VecTrained;

pub use parametric_tsne::ParametricTSNE;

pub use parametric_tsne::ParametricTsneTrained;

pub use random_walk_embedding::RandomWalkEmbedding;

pub use random_walk_embedding::RandomWalkEmbeddingTrained;

pub use sne::SneTrained;

pub use sne::SNE;

pub use spectral_embedding::SpectralEmbedding;

pub use spectral_embedding::SpectralEmbeddingTrained;

pub use symmetric_sne::SymmetricSNE;

pub use symmetric_sne::SymmetricSneTrained;

pub use quality_metrics::*;

pub use stress_testing::*;

pub use geodesic_distance::*;

pub use diffusion_distance::*;

pub use riemannian::*;

pub use topological::*;

pub use distance_kernels::*;

pub use graph_neural_networks::*;

pub use random_projections::*;

pub use similarity::*;

pub use hierarchical::*;

pub use temporal::*;

pub use robust::*;

pub use multi_view::*;

pub use nystrom::*;

pub use compressed_sensing::*;

pub use parallel_knn::*;

pub use stochastic::*;

pub use manifold_traits::*;

pub use fluent_api::*;

pub use extensible_metrics::*;

pub use type_safe_geometry::*;

pub use visualization::*;

pub use plugin_architecture::*;

pub use information_theory::*;

pub use optimal_transport::*;

pub use iterative_refinement::*;

pub use pipeline_middleware::*;

pub use embedding_callbacks::*;

pub use category_theory::*;

pub use performance_optimization::*;

pub use deep_learning::*;

pub use computer_vision::*;

pub use adversarial::*;

pub use continuous_normalizing_flows::*;

pub use nlp::*;

pub use quantum::*;

pub use causal::*;

pub use bioinformatics::*;

Modules

adversarial

Adversarial manifold learning module Adversarial Manifold Learning implementation This module provides adversarial manifold learning methods that use adversarial training to learn robust manifold representations. These methods are particularly effective for learning manifolds that are resilient to adversarial perturbations and noise.

benchmark_datasets

Benchmark datasets module Standard benchmark datasets for manifold learning evaluation This module provides implementations of widely-used benchmark datasets in manifold learning research, enabling fair comparison with reference implementations and standardized evaluation of algorithm performance.

bioinformatics

Bioinformatics applications for manifold learning including genomic analysis, protein structures, phylogenetics, single-cell trajectories, and metabolic pathways

category_theory

Category theory-based manifold representations and functorial embeddings

causal

Causal inference on manifolds module Causal Inference on Manifolds

compressed_sensing

Compressed sensing module Compressed Sensing on Manifolds

computer_vision

Computer vision applications for manifold learning including image patch embedding and face analysis

condition_monitoring

Condition number monitoring for numerical stability Condition number monitoring for manifold learning This module provides comprehensive condition number analysis and monitoring for matrices and operations in manifold learning algorithms. Helps detect numerical instabilities and ill-conditioned problems.

continuous_normalizing_flows

Continuous normalizing flows module Continuous Normalizing Flows for Manifold Learning This module provides implementations of Continuous Normalizing Flows (CNFs), which are a powerful class of generative models that learn invertible transformations between probability distributions using neural ordinary differential equations (NODEs).

deep_learning

Deep learning integration for manifold learning including autoencoders and variational autoencoders

deepwalk

DeepWalk algorithm module DeepWalk Algorithm implementation This module provides DeepWalk for learning continuous representations of vertices in graphs.

dictionary_learning

Dictionary Learning module Dictionary Learning implementation This module provides Dictionary Learning for manifold learning through sparse representation.

diffusion_distance

Diffusion distance computation for manifold analysis

diffusion_maps

Diffusion Maps module Diffusion Maps implementation

distance_kernels

Distance methods and kernel functions module Distance methods and kernel functions for manifold learning

embedding_callbacks

Embedding callbacks for monitoring and customizing manifold learning training Embedding callbacks for monitoring and customizing manifold learning

extensible_metrics

Extensible distance metrics registry Extensible distance metrics for manifold learning

fluent_api

Fluent API for manifold learning configuration Fluent API for manifold learning configuration and execution

geodesic_distance

Geodesic distance computation for manifold-aware metrics

graph_neural_networks

Graph Neural Networks module Graph Neural Network implementations for manifold learning This module provides Graph Convolutional Networks (GCN), GraphSAGE, and Graph Attention Networks (GAT) for learning node embeddings on graph-structured data.

heavy_tailed_symmetric_sne

HeavyTailedSymmetricSNE (Heavy-Tailed Symmetric SNE) module Heavy-Tailed Symmetric SNE implementation This module provides Heavy-Tailed Symmetric SNE for non-linear dimensionality reduction with configurable Student-t distribution.

hessian_lle

HLLE (Hessian LLE) module Hessian Locally Linear Embedding (HLLE) implementation

hierarchical

Hierarchical manifold learning module Hierarchical manifold learning methods This module implements various hierarchical approaches for manifold learning, including multi-scale embeddings and coarse-to-fine optimization.

information_theory

Information-theoretic manifold learning methods Information-theoretic manifold learning methods

isomap

Isomap (Isometric Mapping) module Isomap (Isometric Mapping) implementation

iterative_refinement

Iterative refinement methods for improved numerical stability Iterative refinement methods for improved numerical stability

laplacian_eigenmaps

Laplacian Eigenmaps module Laplacian Eigenmaps implementation

lle

LLE (Locally Linear Embedding) module Locally Linear Embedding (LLE) implementation

ltsa

LTSA (Local Tangent Space Alignment) module Local Tangent Space Alignment (LTSA) implementation

manifold_traits

Trait-based manifold learning framework Trait-based manifold learning framework

mds

MDS (Multidimensional Scaling) module Classical Multidimensional Scaling (MDS) implementation

memory_profiler

Memory profiler module Memory profiling utilities for manifold learning algorithms

minibatch_tsne

Mini-batch t-SNE module Mini-batch t-SNE implementation This module provides Mini-batch t-SNE for large-scale datasets that don’t fit in memory.

minibatch_umap

Mini-batch UMAP module Mini-batch UMAP implementation This module provides Mini-batch UMAP for large-scale datasets that don’t fit in memory.

multi_view

Multi-view learning module Multi-View Learning Framework

mvu

MVU (Maximum Variance Unfolding) module Maximum Variance Unfolding (MVU) implementation

nlp

Natural Language Processing manifold learning module Natural Language Processing Manifold Learning

node2vec

Node2Vec algorithm module Node2Vec Algorithm implementation

nystrom

Nyström approximation module Nyström Approximation for Kernel Methods This module provides efficient approximation methods for kernel matrices using the Nyström method, which enables scalable kernel-based manifold learning algorithms by approximating large kernel matrices with smaller submatrices.

optimal_transport

Optimal transport methods for manifold learning Optimal transport methods for manifold learning

parallel_knn

Parallel k-nearest neighbors module Parallel K-Nearest Neighbors Search

parametric_tsne

ParametricTSNE (Parametric t-SNE) module Parametric t-SNE implementation This module provides Parametric t-SNE for non-linear dimensionality reduction with neural network mapping.

performance_optimization

Advanced performance optimizations including cache-friendly data layouts and unsafe optimizations

pipeline_middleware

Pipeline middleware system for composable manifold learning Middleware system for manifold learning pipelines

plugin_architecture

Plugin architecture for custom manifold learning methods Plugin architecture for custom manifold learning methods

quality_metrics

Quality metrics for evaluating manifold learning algorithms

quantum

Quantum methods for manifold learning module Quantum Methods for Manifold Learning

random_projections

Random projection methods module Random projection methods for high-dimensional data reduction This module implements Johnson-Lindenstrauss embeddings and various random projection techniques for efficient dimensionality reduction with theoretical guarantees.

random_walk_embedding

Random Walk Embedding module Random Walk Embedding implementation

reference_tests

Comparison tests against reference implementations Comparison tests against reference implementations This module provides comprehensive testing infrastructure to compare our manifold learning implementations against reference implementations from scikit-learn and other established libraries.

riemannian

Riemannian manifold processing algorithms

robust

Robust manifold learning module Robust manifold learning methods This module implements robust approaches for manifold learning that can handle outliers, noise, and contamination in the data while preserving the underlying manifold structure.

robust_optimization

Robust optimization methods for manifold learning Robust optimization methods for manifold learning This module provides optimization algorithms that are resistant to numerical instabilities, outliers, and poor conditioning. These methods are essential for reliable manifold learning in real-world applications.

simd_distance

SIMD-optimized distance computations SIMD-optimized distance computations for manifold learning

similarity

Similarity learning module Similarity learning methods for manifold learning This module implements various similarity learning approaches for manifold embeddings, including metric learning, contrastive learning, and related techniques.

sne

SNE (Stochastic Neighbor Embedding) module Stochastic Neighbor Embedding (SNE) implementation This module provides SNE for non-linear dimensionality reduction through probabilistic neighbor embedding.

spectral_embedding

Spectral Embedding module Spectral Embedding implementation

stable_eigenvalue

Numerically stable eigenvalue algorithms Numerically stable eigenvalue algorithms

stochastic

Stochastic manifold learning module Stochastic Manifold Learning Algorithms This module provides stochastic manifold learning algorithms designed for large datasets that cannot fit in memory or require online/streaming processing. These algorithms use random sampling, mini-batch processing, and incremental updates to handle massive datasets efficiently.

stress_testing

Stress testing utilities for manifold learning algorithms

symmetric_sne

SymmetricSNE (Symmetric Stochastic Neighbor Embedding) module Symmetric Stochastic Neighbor Embedding (Symmetric SNE) implementation This module provides Symmetric SNE for non-linear dimensionality reduction through symmetric probabilistic neighbor embedding.

temporal

Temporal manifold learning module Temporal manifold learning methods

timing_utilities

Timing utilities module Timing utilities for embedding speed benchmarks

topological

Topological Data Analysis (TDA) algorithms

tsne

t-SNE (t-distributed Stochastic Neighbor Embedding) module t-SNE (t-distributed Stochastic Neighbor Embedding) implementation

type_safe_geometry

Type-safe geometric operations with compile-time dimension checking Type-safe geometric operations for manifold learning

type_safe_manifolds

Type-safe manifold abstractions with phantom types Type-safe manifold abstractions with phantom types

umap

UMAP module Uniform Manifold Approximation and Projection (UMAP) implementation This module provides UMAP for non-linear dimensionality reduction through uniform manifold approximation.

validation

Validation framework for hyperparameter tuning Validation framework for manifold learning algorithms This module provides comprehensive validation and hyperparameter tuning capabilities for manifold learning algorithms, including cross-validation, grid search, random search, and advanced hyperparameter optimization.

visualization

Visualization integration utilities Visualization integration utilities for manifold learning

zero_cost_abstractions

Zero-cost abstractions for manifold learning Zero-cost abstractions for manifold learning

Macros

time_block

Macro for convenient timing of code blocks

track_array_allocation

Macro for automatic memory tracking of array operations

Structs

SCTrained

SparseCoding

Sparse Coding for manifold learning