Struct LinearAlgebraLibrary 

pub struct LinearAlgebraLibrary {
    pub matrix_storage: MatrixStorage,
    pub computation_engine: ComputationEngine,
    pub optimization_engine: OptimizationEngine,
    pub privacy_engine: PrivacyEngine,
    pub performance_monitor: LAPerformanceMonitor,
}

Linear Algebra Library Manager

Fields

matrix_storage: MatrixStorage

computation_engine: ComputationEngine

optimization_engine: OptimizationEngine

privacy_engine: PrivacyEngine

performance_monitor: LAPerformanceMonitor

Implementations

impl LinearAlgebraLibrary

pub fn new() -> Self

Create new linear algebra library

pub fn initialize(&mut self) -> Result<(), LinearAlgebraError>

Initialize the library

pub fn create_matrix( &mut self, matrix_id: String, rows: usize, cols: usize, data_type: DataType, data: Vec<f64>, ) -> Result<Matrix, LinearAlgebraError>

Create a new matrix

pub fn matrix_multiply( &mut self, left_id: &str, right_id: &str, result_id: &str, alpha: f64, beta: f64, ) -> Result<LinearAlgebraResult<Matrix>, LinearAlgebraError>

Matrix multiplication with hardware acceleration

pub fn matrix_add( &mut self, left_id: &str, right_id: &str, result_id: &str, alpha: f64, ) -> Result<LinearAlgebraResult<Matrix>, LinearAlgebraError>

Matrix addition

pub fn matrix_transpose( &mut self, input_id: &str, result_id: &str, ) -> Result<LinearAlgebraResult<Matrix>, LinearAlgebraError>

Matrix transpose

pub fn matrix_inverse( &mut self, input_id: &str, result_id: &str, ) -> Result<LinearAlgebraResult<Matrix>, LinearAlgebraError>

Matrix inverse

pub fn solve_linear_system( &mut self, matrix_id: &str, rhs_id: &str, solution_id: &str, ) -> Result<LinearAlgebraResult<Matrix>, LinearAlgebraError>

Solve linear system Ax = b

pub fn private_matrix_multiply( &mut self, left_id: &str, right_id: &str, result_id: &str, ) -> Result<LinearAlgebraResult<Matrix>, LinearAlgebraError>

Privacy-preserving matrix multiplication Multiply two matrices and produce a zero-knowledge proof that the published result really is A·B, WITHOUT revealing A or B.

The proof is over a real R1CS circuit (see ZkProofSystem::prove_matrix_multiply): the A/B entries are private witnesses, the result entries are public inputs, and the circuit enforces Σ_k A[i][k]·B[k][j] = C[i][j]. privacy_preserved is set only when that Groth16 proof actually verifies — it is now a genuine cryptographic attestation, not a structural check.

The ZK circuit operates over a FIXED-POINT encoding: each entry is scaled by 1e6 and rounded to a field integer, so real-valued matrices are supported to ~1e-6 precision (integer matrices are encoded exactly). The proof attests the exact scaled-integer identity; the returned matrix is that result rescaled.

pub fn analyze_matrix( &mut self, matrix_id: &str, ) -> Result<MatrixAnalysis, LinearAlgebraError>

Analyze a matrix: detect structural patterns and return optimization hints. Uses the MatrixAnalyzer to detect diagonal, triangular, symmetric, sparse, banded, block-diagonal, Toeplitz, orthogonal, circulant, Hankel, and positive-definite patterns.

pub fn get_performance_stats(&self) -> SystemMetrics

Get performance statistics

pub fn performance_summary(&self) -> String

Get a human-readable performance summary

pub fn cache_hit_rate(&self) -> f64

Get the cache hit rate (0.0 to 1.0)

pub fn cache_size(&self) -> usize

Get the current cache size in bytes

pub fn list_matrices(&self) -> Vec<String>

List all matrices

pub fn get_matrix_info(&self, matrix_id: &str) -> Option<MatrixMetadata>

Get matrix information