Pedagogy: create 60_linear_algebra_2/230_QR_Method_Eigenvalues.ipynb with detailed explanations

[kwlee2025cpp]

Add a new notebook 60_linear_algebra_2/230_QR_Method_Eigenvalues.ipynb between Power Method (200) and Jacobi (240), covering the QR algorithm for general (possibly non-symmetric, possibly complex-eigenvalued) matrices.

Code scope (Scope B from IMPROVEMENTS.md)

• Basic QR iteration: Aₖ₊₁ = Rₖ Qₖ with shift-free demonstration.
• Real Schur form with 2×2 diagonal blocks for complex conjugate eigenpairs.
• Slider over iteration step + Hinton diagram of Aₖ showing the upper-triangularization.

Explanation scope (the deliverable that gates the merge)

The advanced-topic-policy gating criterion is teachability. Author the following:

• Why QR works — informal: each iteration is a similarity transformation that systematically pushes mass into the upper triangle. Connection to Power Method — QR is "subspace iteration" applied to all columns at once.
• Why complex eigenvalues become 2×2 real blocks — a real matrix cannot have a single real diagonal entry equal to a complex eigenvalue; the conjugate pair lives as the eigenvalues of a 2×2 sub-block.
• What "real Schur form" means vs the complex Schur form, and why textbooks usually show the real version.
• Convergence behavior — geometric rate driven by eigenvalue ratio; why a shift accelerates it (mention without implementing if Scope B).
• Where this sits in the curriculum — reading Power Method (200) as "find one eigenvector" and Jacobi (240) as "diagonalize a symmetric matrix"; QR fills the gap of "find all eigenvalues of a general matrix".

Note on policy

This is a brand-new notebook on an advanced topic. Per feedback_nmisp_teach_what_you_know.md, brand-new advanced material is mergeable only if the explanatory scaffolding is part of the same PR — not deferred to "we'll write the prose later."