1: \begin{abstract}

2: Some important applicative problems require the evaluation of

3: functions $\Psi$ of large and sparse and/or \emph{localized}

4: matrices $A$. Popular and interesting techniques for computing

5: $\Psi(A)$ and $\Psi(A)\mathbf{v}$, where $\mathbf{v}$ is a vector,

6: are based on partial fraction expansions. However, some of these

7: techniques require solving several linear systems whose matrices

8: differ from $A$ by a complex multiple of the identity matrix $I$ for

9: computing $\Psi(A)\mathbf{v}$ or require inverting sequences of

10: matrices with the same characteristics for computing $\Psi(A)$. Here

11: we study the use and the convergence of a recent technique for

12: generating sequences of incomplete factorizations of matrices in

13: order to face with both these issues. The solution of the sequences

14: of linear systems and approximate matrix inversions above can be

15: computed efficiently provided that $A^{-1}$ shows certain decay

16: properties. These strategies have good parallel potentialities.

17: 

18: Our claims are confirmed by numerical tests.

19: \end{abstract}