cherab.phix.inversion.GCV.eta#

GCV.eta(beta)Source#

Calculate squared regularization norm: \(\eta = ||Lx_\lambda||^2\)

\(\eta\) can be calculated with SVD components as follows:

\[\begin{split}\eta &= \left\| V W \Sigma^{-1} U^\mathsf{T} b \right\|^2\\ &= \left\| \begin{pmatrix} v_1 & \cdots & v_r \end{pmatrix} W \Sigma^{-1} U^\mathsf{T} b \right\|^2\\ &= \left\| W \Sigma^{-1} U^\mathsf{T} b \right\|^2 \quad( \because V^\mathsf{T}V = I_r, \quad\text{i.e.}\quad v_i\cdot v_j = \delta_{ij} ),\end{split}\]

where \(W = \text{diag}(w_1(\lambda), ..., w_r(\lambda))\) and \(w_i(\lambda)\) is the window function.

Parameters:

beta (float) – regularization parameter

Returns:

squared regularization norm

Return type:

numpy.floating