Inverse iteration (source code)

= Inverse iteration
{wiki}

Inverse iteration repeatedly solves $(A-sI)y_{k+1}=q_k$ and applies <vector normalization>. It is the <power method> for $(A-sI)^{-1}$, so $s$ must not be an <eigenvalue>. For a real <symmetric matrix>, if one <eigenvalue> is uniquely closest to $s$ and the starting <vector> has a nonzero component in its <eigenspace>, the normalized directions converge to that <eigenspace>. Ties or an absent initial component can prevent that conclusion.