pypcga.PCGA.get_invA_as_linop#

PCGA.get_invA_as_linop(HZ: ndarray[tuple[Any, ...], dtype[float64]], HX: ndarray[tuple[Any, ...], dtype[float64]], cov_obs: CovarianceMatrix) → InvALinOp[source]#

Return the low rank inverse of A as a linear operator.

The output is used as a preconditioner for Krylov subspace iterative approaches when solving Ax = b.

See section 2.3 in Lee et al. [2016].