PolarDecomposition<T>

Models & Types ยท AiDotNet.DecompositionMethods.MatrixDecomposition

Implements the Polar Decomposition of a matrix, which factors a matrix A into the product of an orthogonal matrix U and a positive semi-definite matrix P.

Inheritance: MatrixDecompositionBase<T> โ†’ PolarDecomposition<T>

For Beginners

Think of Polar Decomposition as breaking down a transformation into two simpler steps: first a rotation/reflection (U), and then a stretching/scaling (P). This is similar to how polar coordinates break down a point into an angle and a distance.

How It Works

The Polar Decomposition expresses a matrix A as A = UP, where U is orthogonal (U^T * U = I) and P is positive semi-definite.

Constructors

Constructor Summary
PolarDecomposition(Matrix<T> matrix, PolarAlgorithmType algorithm) Initializes a new instance of the PolarDecomposition class.

Properties

Property Summary
Matrix<T> P { get; set; } Gets the positive semi-definite factor of the decomposition.
Matrix<T> U { get; set; } Gets the orthogonal factor of the decomposition.

Methods

Method Summary
Matrix<T> Invert() Computes the inverse of the original matrix A using its polar decomposition.
ValueTuple<Matrix<T>, Matrix<T>> GetFactors() Returns the factors U and P of the polar decomposition.
Vector<T> Solve(Vector<T> b) Solves the linear system Ax = b using the polar decomposition.