On Perron-Frobenius property of matrices having some negative entries (Q2576214): Difference between revisions
From MaRDI portal
Created a new Item |
Added link to MaRDI item. |
||
links / mardi / name | links / mardi / name | ||
Revision as of 07:22, 3 February 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On Perron-Frobenius property of matrices having some negative entries |
scientific article |
Statements
On Perron-Frobenius property of matrices having some negative entries (English)
0 references
27 December 2005
0 references
This paper deals with Perron-Frobenius property of matrices which have some negative entries. It is easy to see, from the continuity of the eigenvalues and the entries of the eigenvectors, as functions of the entries of matrices, that the Perron-Frobenius theory may hold also in the case where the matrix has some absolutely small negative entries. In this work, the author analyzes the following questions: How small could these entries be? What is their distribution? When does a change in a matrix with the property of having a Perron eigenvector imply a loss of this property? The author studies matrices that have some negative entries. Specifically, he presents necessary and sufficient conditions for a matrix to have a Perron eigenvector. So, he answers implicitly the above questions by extending the Perron-Frobenius theory of nonnegative matrices to the class of matrices that possess the Perron-Frobenius property. Finally, the author applies this theory by introducing the Perron-Frobenius splitting for the solution of linear systems by classical iterative methods. This splitting is an extension of the well-known regular, weak regular and nonnegative splittings. He also presents properties of convergence and comparison properties for the proposed splitting.
0 references
Perron-Frobenius property
0 references
nonnegative matrices
0 references
Perron-Frobenius splitting
0 references
matrices with some negative entries
0 references
Perron eigenvector
0 references
iterative methods
0 references
convergence
0 references