demo_eigenvalues.py

This demo computes a dominant eigenvalue using power iteration implemented with Tensor operations.

Run

python demos/demo_eigenvalues.py --matrix symmetric3 --iters 200

CLI options

Option	Default	Description
--matrix	symmetric3	Matrix preset (symmetric3, spd4)
--iters	200	Maximum iteration count
--tol	1e-8	Convergence tolerance on eigenvalue change
--seed	42	Random seed for initial vector

Output

The script prints:

• Chosen matrix preset and matrix values
• Iteration count used
• Estimated dominant eigenvalue
• Exact dominant eigenvalue from NumPy
• Absolute error
• Residual norm ||Av - lambda*v||