Dynamics of piecewise linear maps and sets of nonnegative matrices (Q1030713): Difference between revisions

The theory of nonnegative matrices has been very well developed since its appearance in papers of Perron and Frobenius. The classical Perron-Frobenius theorem shows that a nonnegative matrix has a nonnegative eigenvector associated with its spectral radius, and if the matrix is irreducible then this nonnegative eigenvector can be chosen strictly positive. Many generalizations of the Perron-Frobenius theory involve homogeneous monotone functions, especially functions of the form \[ g_K(x)=\max_{A\in K}Ax, \] where \(K\) is a finite set of square nonnegative matrices of fixed dimension and by ``max'' we mean component-wise maximum. The author considers maps of similar form, but with``minimum'' instead of ``maximum'': \[ f_K(x)=\min_{A\in K}Ax. \] He transfers known results about properties of \(g_K\) to \(f_K\). In particular he shows existence of nonnegative generalized eigenvectors of \(f_K\), gives necessary and sufficient conditions for the existence of a strictly positive eigenvector of \(f_K\), and studies dynamics of \(f_K\) on the positive cone. He also shows the existence and constructs matrices \(A\) and \(B\), possibly not in \(K\), such that \(f_K^n(v)\sim A^nv\) and \(g_K^n(v)\sim B^nv\) for any strictly positive vector \(v\).