Module spectral

Module spectral 

Source
Expand description

Spectral Methods for Graph Analysis

Chebyshev polynomials and spectral graph theory for efficient diffusion and filtering without eigendecomposition.

§Key Capabilities

  • Chebyshev Graph Filtering: O(Km) filtering where K is polynomial degree
  • Graph Diffusion: Heat kernel approximation via Chebyshev expansion
  • Spectral Clustering: Efficient k-way partitioning
  • Wavelet Transforms: Multi-scale graph analysis

§Integration with Mincut

Spectral methods pair naturally with mincut:

  • Mincut identifies partition boundaries
  • Chebyshev smooths attention within partitions
  • Spectral clustering provides initial segmentation hints

§Mathematical Background

Chebyshev polynomials T_k(x) satisfy:

  • T_0(x) = 1
  • T_1(x) = x
  • T_{k+1}(x) = 2x·T_k(x) - T_{k-1}(x)

This recurrence enables O(K) evaluation of degree-K polynomial filters.

Structs§

ChebyshevExpansion
Chebyshev expansion of a function f(x) ≈ Σ c_k T_k(x)
ChebyshevPolynomial
Chebyshev polynomial of the first kind
ClusteringConfig
Spectral clustering configuration
GraphFilter
Graph filter that applies spectral operations
GraphWavelet
Graph wavelet at specific vertex
ScaledLaplacian
Scaled Laplacian for Chebyshev approximation L_scaled = 2L/λ_max - I (eigenvalues in [-1, 1])
SpectralClustering
Spectral clustering
SpectralFilter
Spectral graph filter using Chebyshev approximation
SpectralWaveletTransform
Spectral Wavelet Transform
WaveletScale
Wavelet scale configuration

Enums§

FilterType
Type of spectral filter
LaplacianNorm
Normalized Laplacian (symmetric or random walk)