Constructing approximate shell-model wavefunctions by eigenvector continuation

Shell-model calculations play a key role in elucidating various properties of nuclei. In general, those studies require huge number to be repeated for parameter calibration and quantifying uncertainties. To reduce the computational burden, we propose new workflow shell-model using method called eigenvector continuation (EC). It enables us efficiently approximate eigenpairs under given Hamiltonian by previously sampled eigenvectors. We demonstrate validity EC as an emulator valence shell-model, including first application electromagnetic transition matrix elements. Furthermore, usage EC: preprocessing, which start Lanczos iterations from eigenvectors, that this can accelerate subsequent research cycles. With aid EC, eigenvectors obtained during optimization are not necessarily discarded, even if their eigenvalues far experimental data. Those become accumulated knowledge.