<aside> 🔖

Contents

</aside>

Project Description

This project implemented various numerical methods, including the power method, Hermite quadrature, and conjugate gradient method, to solve mathematical problems involving eigenvalue determination, integration of improper integrals, and solving linear systems. The key objective was to calculate eigenvalues, integrate an improper integral using Hermite quadrature, and solve a linear system using the conjugate gradient method, demonstrating the application of these techniques in solving complex problems that cannot be addressed analytically.

Methodology

The power method is used to find the largest eigenvalue and the corresponding eigenvector of a matrix. The eigenvector $x_k$ is computed iteratively as:

$$ x_k = Ax_(k-1) / ||Ax_(k-1)|| $$

where $A$ is the matrix and

$∣∣⋅∣∣$ represents the norm. The eigenvalue $λ$ is calculated as:

$$ λ = <Ax_k, x_k> $$

The inverse shifted power method was also used to find the smallest eigenvalue by transforming the matrix with a shifting parameter μ:

$$ z_k^{-1} = (A - mu I)^{-1}x_(k-1) / ||(A - mu I)^{-1}x_(k-1)|| $$

The eigenvalue is related to the shifted value $z$ by:

$$ z = (λ- μ)^{-1} $$

Hermite quadrature is used to evaluate improper integrals. The approximation formula is given as:

$$ integral( f(x) / sqrt(1 - x^2) ) ≈ (π/ n) * sum( f(x_i) ) $$

where $f(x)$ is the function to be integrated and

$x_i$ are the Gauss-Hermite quadrature points.

The conjugate gradient method solves linear systems of the form $Ax=b$. The iterative formula for updating the solution is:

$$ x_(k+1) = x_k + t_k * v_k $$

where $t_k$ is calculated as: