1: \begin{abstract}

2: In this paper, based on the Noda iteration, we present inexact Noda

3: iterations (INI), to find the smallest eigenvalue and the associated

4: positive eigenvector of a large irreducible nonsingular $M$-matrix. The

5: positivity of approximations is critical in applications, and if the

6: approximations lose the positivity then they may be meaningless and could

7: not be interpreted. We propose two different inner tolerance strategies for

8: solving the inner linear systems involved, and prove that the convergence of

9: resulting INI algorithms is globally linear and superlinear with the

10: convergence order $\frac{1+\sqrt{5}}{2}$, respectively. The proposed INI

11: algorithms are structure preserving and maintains the positivity of

12: approximate eigenvectors. We also revisit the exact Noda iteration and

13: establish a new quadratic convergence result. All the above is first done

14: for the problem of computing the Perron root and the positive Perron vector

15: of an irreducible nonnegative matrix and is then adapted to computing the

16: smallest eigenpair of the irreducible nonsingular $M$-matrix. Numerical

17: examples illustrate that the proposed INI algorithms are practical, and they

18: always preserve the positivity of approximate eigenvectors. We compare them

19: with the Jacobi--Davidson method, the implicitly restarted Arnoldi method

20: and the explicitly restarted Krylov--Schur method, all of which cannot

21: guarantee the positivity of approximate eigenvectors, and illustrate that

22: the overall efficiency of the INI algorithms is competitive with and can be

23: considerably higher than the latter three methods.

24: 

25: % \PACS{PACS code1 \and PACS code2 \and more}

26: \subclass{Primary 15A18 \and 65F15 \and 65F50; Secondary 15B48 \and 15B99.}

27: %\subclass{MSC code1 \and MSC code2 \and more}

28: \end{abstract}