Su un programma di calcolo per la determinazione di alcuni autovalori di una matrice simmetrica secondo il metodo di Davidson
A computer implementation of E. R. Davidson's algorithm for iterative evaluation of the few lowest eigenvalues and eigenvectors of large (symmetric) matrices is presented. The performance of the method is tested on a 61x61 matrix with satisfactory results regarding both convergence rate and numerical precision. Comparisons are made with results obtained by Givens and Nesbet-deflation methods.