Expand description
Covariance estimation algorithms
This module provides covariance estimators including empirical, shrinkage-based, and robust estimators.
Re-exports§
pub use plugin_architecture::GLOBAL_REGISTRY;
Modules§
- model_
selection_ presets - Convenience functions for creating common selectors
- polars_
utils - Utility functions for DataFrame integration
- tuning_
presets - Convenience functions for creating common tuning configurations
Macros§
- register_
covariance_ estimator - Convenience macros for plugin registration
- register_
hook - register_
regularization
Structs§
- ALSCovariance
- Alternating Least Squares estimator for covariance matrices
- ALSCovariance
Trained - Trained ALS state
- Accuracy
Test Result - Adam
Optimizer - Adam optimizer
- Adaptive
Filtering Covariance - Adaptive filtering covariance
- Adaptive
Filtering Covariance Trained - Trained state for adaptive filtering covariance
- Adaptive
Filtering Covariance Untrained - Untrained state for adaptive filtering covariance
- Adaptive
Lasso Covariance - Adaptive Lasso Covariance Estimator
- Adaptive
Lasso Covariance Trained - Trained state for AdaptiveLassoCovariance
- Adversarial
Robust Covariance - Adversarially robust covariance estimator
- Adversarial
Robust Covariance Trained - Adversarial
Robust Covariance Untrained - States for adversarial robust covariance
- Alternating
Projections - Alternating Projections Estimator (Untrained State)
- Alternating
Projections Trained - Marker for trained state
- Array
Signal Processing - Array signal processing covariance
- Array
Signal Processing Trained - Trained state for array signal processing
- Array
Signal Processing Untrained - Untrained state for array signal processing
- Auto
Covariance Selector - Automatic model selection framework for covariance estimators
- Bayesian
Covariance - Bayesian covariance estimator in untrained state
- Bayesian
Covariance Fitted - Bayesian covariance estimator in trained state
- Bayesian
Prior - Prior specification for Bayesian covariance estimation
- Beamforming
Covariance - Beamforming covariance applications
- Beamforming
Covariance Trained - Trained state for beamforming covariance
- Beamforming
Covariance Untrained - Untrained state for beamforming covariance
- Benchmark
Config - Benchmark
Result - Results from performance benchmarking
- Benchmark
Suite - Best
Estimator - Best estimator information
- BigQUIC
- BigQUIC estimator for large-scale sparse precision matrix estimation
- BigQUIC
Trained - Trained BigQUIC state
- CLIME
- CLIME (Constrained L1 Minimization) estimator
- CVResult
- Single cross-validation result
- Candidate
Result - Individual candidate result
- Chen
Stein Covariance - Chen-Stein Shrinkage Covariance Estimator
- Chen
Stein Covariance Trained - Trained state for ChenSteinCovariance
- Clutter
Statistics - Clutter statistics
- Column
Statistics - Statistical summary for a column
- Communication
Cost - Communication cost analysis
- Comparison
Result - Composite
Regularization - Composite regularization strategy
- Computation
Stats - Computational
Constraints - Computational constraints
- Conditioning
Step - Matrix conditioning post-processing step
- Convergence
Analysis - Convergence analysis
- Convergence
Info - Convergence information for iterative estimators
- Convergence
Params - Convergence parameters
- Coordinate
Descent Covariance - Coordinate descent covariance estimator
- Coordinate
Descent Covariance Trained - Trained Coordinate Descent state
- Coordinate
Descent Optimizer - Coordinate descent optimizer
- Copula
Config - Copula-based estimation configuration
- Correlation
Diagnostics - Correlation-based diagnostics
- Correlation
Structure - Correlation structure analysis
- Covariance
Accumulator - Covariance
Benchmark - Performance benchmarking utilities for covariance estimators
- CovarianceCV
- Cross-validation utility for covariance estimator selection
- Covariance
Data Frame - Polars DataFrame wrapper for covariance estimation
- Covariance
Diagnostics - Comprehensive diagnostic report for a covariance matrix
- Covariance
Hyperparameter Tuner - Hyperparameter tuner for covariance estimators
- Covariance
Iterator - Covariance
Metadata - Covariance
Pipeline - Fluent covariance estimation pipeline builder
- Covariance
Plugin Registry - Plugin registry for custom covariance estimators
- Covariance
Presets - Configuration presets for common covariance estimation scenarios
- Covariance
Properties - Properties of a covariance matrix
- Covariance
Result - Result of covariance estimation with DataFrame context
- Cross
Validation Config - Cross-validation configuration
- Data
Characteristics - Data characteristics analysis
- Data
Characterization Rules - Data characterization rules
- Data
Frame Description - Summary description of a DataFrame
- Data
Frame Metadata - Metadata about the DataFrame columns
- DccConfig
- DCC model specification
- Diagonal
- Diagonal
Stats - Statistics for diagonal elements (variances)
- Differential
Privacy Covariance - Basic Differential Privacy Covariance Estimator (Untrained)
- Distributed
Covariance - Distribution
Characteristics - Distribution characteristics
- Distribution
Free Config - Distribution-free estimation configuration
- Domain
Presets - Domain-specific preset collections
- EMCovariance
Missing Data - EM-based covariance estimator for missing data
- EMCovariance
Missing Data Trained - Trained EM Covariance state
- Early
Stopping Config - Early stopping configuration
- Elastic
NetCovariance - Elastic Net Regularized Covariance Estimator
- Elastic
NetCovariance Trained - Trained state for ElasticNetCovariance
- Elliptic
Envelope - Elliptic Envelope
- Elliptic
Envelope Trained - Trained state for EllipticEnvelope
- Empirical
Covariance - Empirical Covariance Estimator
- Empirical
Covariance Trained - Trained state for EmpiricalCovariance
- Estimator
Config - Estimator configuration options
- Estimator
Info - Information about the estimator used
- Estimator
Metadata - Metadata for custom estimators
- Exploration
Metrics - Metrics about search space exploration
- Exponential
Weighted Config - Exponentially weighted configuration
- Factor
Model Covariance - Factor Model Covariance estimator
- Factor
Model Covariance Trained - Trained Factor Model state
- Federated
Covariance - Federated covariance estimator
- Federated
Covariance Trained - Federated
Covariance Untrained - States for federated covariance
- Federated
Party - Federated party data structure
- Financial
- Financial domain presets
- Fitted
- Fluent
Cross Validation Config - Cross-validation configuration
- Framework
Optimization History - Optimization history tracking
- Frank
Wolfe Covariance - Frank-Wolfe Covariance Estimator (Untrained State)
- Frank
Wolfe Covariance Trained - Marker for trained state
- GLOBAL_
REGISTRY - Garch
Config - GARCH model configuration
- Gene
Expression Network - Gene expression covariance network estimation
- Gene
Expression Network Trained - Trained state for gene expression network
- Gene
Expression Network Untrained - Untrained state for gene expression network
- General
- Generic
Empirical Covariance - Genomics
- Genomics domain presets
- Graphical
Lasso - Graphical Lasso Estimator
- Graphical
Lasso Trained - Trained state for GraphicalLasso
- Ground
Truth Test Case - Group
Lasso Covariance - Group Lasso Covariance Estimator
- Group
Lasso Covariance Trained - Trained state for GroupLassoCovariance
- Group
Lasso Regularization - Group Lasso regularization
- Heuristic
Rule - Heuristic rules for model selection
- Hook
Context - Context passed to hooks
- Huber
Covariance - Huber Robust Covariance Estimator
- Huber
Covariance Trained - Trained state for HuberCovariance
- ICACovariance
- ICA-based covariance estimator
- ICACovariance
Trained - Trained ICA Covariance state
- IPFCovariance
- Iterative Proportional Fitting covariance estimator
- IPFCovariance
Trained - Trained IPF Covariance state
- Information
Metrics - Information geometry metrics
- Information
Theory Covariance - Information Theory Covariance Estimator
- KdeConfig
- Kernel density estimation configuration
- L1Regularization
- L1 regularization (Lasso)
- L2Regularization
- L2 regularization (Ridge)
- Ledoit
Wolf - Ledoit-Wolf Covariance Estimator
- Ledoit
Wolf Trained - Trained state for LedoitWolf
- LowRank
Sparse Covariance - Low-Rank Plus Sparse decomposition estimator for covariance matrices
- LowRank
Sparse Covariance Trained - Trained Low-Rank Sparse state
- Marginal
Constraint - Marginal constraint specification
- Mcmc
Config - MCMC sampling configuration
- Memory
Efficient Covariance - Memory
Estimate - Meta
Features - Meta-features extracted from datasets
- Meta
Learning Covariance - Meta-Learning Covariance Estimator
- Meta
Performance Metrics - Performance metrics for meta-learning
- MinCov
Det - Minimum Covariance Determinant (MCD) Estimator
- MinCov
DetTrained - Trained state for MinCovDet
- Missing
Data Info - Missing data information
- Model
FitStatistics - Model fit statistics
- Model
SelectionCV - Cross-validation configuration for model selection
- Model
Selection Result - Model selection result
- Multi
Omics Covariance - Multi-omics covariance estimation
- Multi
Omics Covariance Trained - Trained state for multi-omics covariance
- Multi
Omics Covariance Untrained - Untrained state for multi-omics covariance
- NMFCovariance
- NMF Covariance Estimator (Untrained State)
- NMFCovariance
Trained - Marker for trained state
- Neighborhood
Selection - Neighborhood Selection Estimator
- Neighborhood
Selection Trained - Trained state for NeighborhoodSelection
- Nelder
Mead Optimizer - Nelder-Mead simplex optimizer (derivative-free)
- Network
Statistics - Network statistics
- Noise
Characteristics - Noise characteristics
- Nonlinear
Shrinkage - Nonlinear shrinkage covariance estimator
- Nonparametric
Covariance - Non-parametric covariance estimator in untrained state
- Nonparametric
Covariance Fitted - Non-parametric covariance estimator in trained state
- Nuclear
Norm Minimization - Nuclear norm minimization estimator for matrix completion
- Nuclear
Norm Regularization - Nuclear norm regularization
- Numerical
Accuracy Tester - Numerically
Stable Covariance - OAS
- Oracle Approximating Shrinkage (OAS) Estimator
- OASTrained
- Trained state for OAS
- OffDiagonal
Stats - Statistics for off-diagonal elements (covariances)
- Optimization
Config - Optimization configuration
- Optimization
Config Builder - Builder for optimization configuration
- Optimization
History - Optimization history tracking
- Optimization
Result - Optimization result
- Optimizer
Registry - Registry for managing optimization algorithms
- Outlier
Removal Step - Outlier removal preprocessing step
- PCACovariance
- PCA-based covariance estimator
- PCACovariance
Trained - Trained PCA Covariance state
- Parallel
Covariance - Parallel
Covariance Trained - Parallel
Covariance Untrained - Parameter
Spec - Parameter specification for tuning
- Pathway
Analysis - Pathway analysis integration
- Pathway
Analysis Trained - Trained state for pathway analysis
- Pathway
Analysis Untrained - Untrained state for pathway analysis
- Performance
Comparison - Performance comparison statistics
- Performance
History - Performance history for meta-learning
- Performance
Metrics - Performance metrics for the estimation
- Phylogenetic
Covariance - Phylogenetic covariance estimation
- Phylogenetic
Covariance Trained - Trained state for phylogenetic covariance
- Phylogenetic
Covariance Untrained - Untrained state for phylogenetic covariance
- Pipeline
Metadata - Pipeline metadata
- Plugin
Parameter Spec - Parameter specification
- Portfolio
Optimizer - Positive
Definite - Preset
Recommendations - Preset validation and recommendations
- Privacy
Accountant - Privacy accounting for composition
- Privacy
Operation - Individual privacy operation
- Property
Failure - Property
Test Result - Property
Tester - Protein
Interaction Network - Protein interaction network estimation
- Protein
Interaction Network Trained - Trained state for protein interaction network
- Protein
Interaction Network Untrained - Untrained state for protein interaction network
- Proximal
Gradient Optimizer - Proximal gradient optimizer
- Quantum
Advantage Analysis - Analysis of quantum computational advantage
- Quantum
Inspired Covariance - Quantum-inspired covariance estimation using quantum algorithms principles
- Quantum
Inspired Covariance Trained - Quantum
Inspired Covariance Untrained - States for quantum-inspired covariance estimation
- Radar
Sonar Covariance - Radar and sonar covariance applications
- Radar
Sonar Covariance Trained - Trained state for radar/sonar covariance
- Radar
Sonar Covariance Untrained - Untrained state for radar/sonar covariance
- Rank
Based Config - Rank-based estimation configuration
- RaoBlackwell
Ledoit Wolf - Rao-Blackwell Ledoit-Wolf Covariance Estimator
- RaoBlackwell
Ledoit Wolf Trained - Trained state for RaoBlackwellLedoitWolf
- Regime
Switching Config - Regime-switching configuration
- Regularization
Factory - Regularization factory for creating common combinations
- Ridge
Covariance - Ridge Regularized Covariance Estimator
- Ridge
Covariance Trained - Trained state for RidgeCovariance
- Risk
Decomposition - Risk
Factor Model - Risk
Factor Model Trained - Risk
Factor Model Untrained - Robust
Correlation Config - Robust correlation configuration
- RobustPCA
- Robust Principal Component Analysis (RPCA) Estimator
- RobustPCA
Trained - Trained state for RobustPCA
- Robustness
Diagnostics - Robustness diagnostics
- Rolling
Window Config - Rolling window configuration
- Rotation
Equivariant - Rotation-Equivariant Shrinkage Covariance Estimator
- Rotation
Equivariant Trained - Trained state for RotationEquivariant
- SGDOptimizer
- Stochastic Gradient Descent optimizer
- SIMD
Covariance - SPACE
- SPACE (Sparse Partial Correlation Estimation) Estimator
- SPACE
Trained - Trained state for SPACE
- Shared
Covariance - Shrunk
Covariance - Shrunk Covariance Estimator
- Shrunk
Covariance Trained - Trained state for ShrunkCovariance
- Signal
Processing - Signal processing domain presets
- Single
Accuracy Result - Single
Comparison - Sparse
Factor Model - Sparse Factor Model (Untrained State)
- Sparse
Factor Model Trained - Marker for trained state
- Spatial
Covariance Estimator - Spatial covariance estimation for array processing
- Spatial
Covariance Estimator Trained - Trained state for spatial covariance estimator
- Spatial
Covariance Estimator Untrained - Untrained state for spatial covariance estimator
- Standardization
Step - Standardization preprocessing step
- Step
Result - Step execution result
- Streaming
Covariance - Stress
Scenario - Stress
Test Result - Stress
Testing - Symmetric
- TIGER
- TIGER (Tuning-Insensitive Graph Estimation) Estimator
- TIGER
Trained - Trained state for TIGER
- Testing
Benchmark Result - Thread
Safe Covariance View - Time
Varying Covariance - Time-varying covariance estimator in untrained state
- Time
Varying Covariance Fitted - Time-varying covariance estimator in trained state
- Topology
Metrics - Network topology metrics
- Tuning
Config - Hyperparameter tuning configuration
- Tuning
Result - Result of hyperparameter tuning
- Typed
Matrix - Unfit
- Pipeline states
- Utility
Metrics - Utility metrics for privacy-utility trade-off
- Variational
Config - Variational Bayes configuration
- Variational
Parameters - Variational parameters for Bayesian inference
- Volatility
Model - Zero
Cost Covariance
Enums§
- APAlgorithm
- Alternating projections algorithm variants
- Acquisition
Function - Acquisition function for Bayesian optimization
- Adaptive
Algorithm - Adaptive algorithms
- Aggregation
Method - Methods for federated aggregation
- Algorithm
Complexity - Algorithm complexity classifications
- Array
Geometry - Array geometry types
- Bayesian
Method - Bayesian covariance estimation methods
- Beamforming
Algorithm - Beamforming algorithms
- Branch
Length Method - Branch length estimation methods
- Budget
Allocation - Privacy budget allocation strategies
- Clustering
Method - Clustering methods for gene network analysis
- Clutter
Suppression - Clutter suppression methods
- Combination
Method - Methods for combining regularization strategies
- Complex
Detection Method - Methods for protein complex detection
- Composition
Method - Composition methods for privacy accounting
- Computational
Complexity - Computational complexity classification
- Conditioning
Method - Constraint
Type - Types of constraints for IPF
- Contrast
Function - Contrast functions for FastICA
- Convergence
Criterion - Convergence criteria
- Copula
Type - Copula types for dependency modeling
- Correction
Method - Correction methods for multiple testing
- Correlation
Handling - Correlation handling methods
- Covariance
Error - Covariance
Method - Available covariance estimation methods for meta-learning
- DOAMethod
- Direction of arrival estimation methods
- Detection
Method - Detection methods
- Difficulty
Level - Distribution
Strategy - Divergence
Measure - Divergence measures for information theory
- Doppler
Processing - Doppler processing methods
- Enrichment
Method - Enrichment analysis methods
- Entropy
Estimator - Entropy estimation methods
- Estimator
State - Estimator state for hooks
- Estimator
Type - Available estimator types
- Evolutionary
Model - Evolutionary models
- Factor
Model Method - Federated
Privacy Mechanism - Privacy mechanisms for federated learning
- Filter
Type - Filter types
- Fluent
Scoring Metric - Scoring metrics for cross-validation
- Frank
Wolfe Algorithm - Frank-Wolfe algorithm variants
- Frank
Wolfe Constraint - Constraint sets for Frank-Wolfe optimization
- Frank
Wolfe Line Search Method - Line search methods for Frank-Wolfe
- Frank
Wolfe Objective Function - Objective functions for covariance optimization
- Garch
Type - GARCH model types for multivariate estimation
- Hook
Type - Hook types for different estimation phases
- ICAAlgorithm
- ICA algorithms
- Information
Criterion - Information criteria for model selection
- Information
Method - Information theory estimation methods
- Information
Regularization - Regularization strategies based on information theory
- Integration
Method - Integration methods for multi-omics data
- Kernel
Function - Kernel functions for kernel PCA
- Kernel
Type - Kernel types for density estimation
- Line
Search Method - Line search methods
- Meta
Learning Strategy - Meta-learning strategy
- Meta
Optimization Method - Hyperparameter optimization method
- Missing
Data Method - Methods for handling missing data
- Model
Selection Scoring - Scoring metrics for model selection
- NMFAlgorithm
- NMF algorithm variants
- NMFInitialization
- NMF initialization methods
- Noise
Calibration - Noise calibration methods
- Noise
Type - Noise types
- Nonparametric
Method - Non-parametric covariance estimation methods
- Normalization
Method - Normalization methods
- Optimization
Method - Optimization
Target - Optimization targets for coordinate descent
- Optimizer
Type - Optimization algorithm types
- Outlier
Method - PCAMethod
- Methods for PCA computation
- Parameter
Type - Types of parameters that can be tuned
- Parameter
Value - Parameter value wrapper
- Performance
Metric - Performance metrics
- Pivoting
Strategy - Privacy
Mechanism - Privacy mechanism types
- Projection
Constraint - Projection constraint types
- Quality
Assessment - Overall quality assessment
- Quantum
Algorithm Type - Types of quantum algorithms for covariance estimation
- Range
Processing - Range processing methods
- Rank
Correlation Type - Rank-based correlation types
- Rank
Method - Methods for handling rank deficiency
- Rate
Variation Model - Rate variation models
- Regularization
Method - Regularization methods for coordinate descent
- Robust
Correlation Type - Robust correlation types
- Robustness
Method - Methods for adversarial robustness
- Rule
Condition - Rule conditions for heuristic selection
- Scoring
Method - Scoring methods for covariance cross-validation
- Scoring
Metric - Scoring metrics for evaluating covariance estimators
- Search
Strategy - Search strategy for hyperparameter optimization
- Selection
Rule - Selection rules for choosing final model
- Selection
Strategy - Model selection strategies
- Sparse
Initialization - Sparse initialization methods
- Sparse
Regularization - Sparse regularization methods
- Spatial
Estimation Method - Spatial estimation methods
- Spatial
Smoothing - Spatial smoothing techniques
- Step
Type - Step types
- Stratification
Strategy - Stratification strategies for CV
- Streaming
Method - System
Type - System types
- Time
Varying Method - Time-varying covariance estimation methods
- Update
Rule - Update rules for multiplicative NMF
- Volatility
Model Type - Whitening
Method - Whitening methods
Traits§
- Covariance
Estimator - Covariance
Estimator Fitted - Trait for fitted covariance estimators
- Covariance
Estimator Tunable - Trait for covariance estimators that can be tuned
- Custom
Covariance Estimator - Trait for custom covariance estimators
- Data
Frame Estimator - Trait for estimators that can work directly with DataFrames
- Estimator
Factory - Factory trait for creating custom estimators
- Hook
- Hook trait for callback functions
- Matrix
Structure - Phantom types for matrix structure
- Middleware
- Middleware trait for estimation pipelines
- Objective
Function - Trait for objective functions
- Optimization
Algorithm - Core trait for optimization algorithms
- Postprocessing
Step - Post-processing step trait
- Preprocessing
Step - Preprocessing step trait
- Regularization
Function - Regularization function trait
- Regularization
Strategy - Trait for regularization strategies
- Robust
Covariance Estimator - Sparse
Covariance Estimator
Functions§
- adaptive_
shrinkage - Stability-oriented covariance shrinkage
- compare_
covariance_ matrices - Compare two covariance matrices
- create_
federated_ parties - Create federated parties from distributed data
- frobenius_
norm - Compute the Frobenius norm of a matrix
- is_
diagonally_ dominant - Check if a matrix is diagonally dominant (useful for convergence analysis)
- matrix_
determinant - Compute matrix determinant
- matrix_
inverse - Compute matrix inverse using appropriate method based on size
- nuclear_
norm_ approximation - Compute the nuclear norm (sum of singular values) approximation via trace
- rank_
estimate - Matrix rank estimation via iterative hard thresholding
- spectral_
radius_ estimate - Estimate the spectral radius (largest eigenvalue magnitude) via power iteration
- split_
data_ for_ federation - Split data for federated simulation
- validate_
covariance_ matrix - Validate properties of a covariance matrix