A note on the eigenvalues of a primitive matrix with large exponent (Q677125)

An \(n\times n\) matrix \(A=(a_{ij})\) with all \(a_{ij}\geq 0\) is called primitive if \(A^k\) has all positive entries for some \(k\geq 1\); the minimum such \(k\) is called \(\exp(A)\), the exponent of \(A\). A matrix \(A=(a_{ij})\) is stochastic, if all \(a_{ij}\geq 0\) and \(\sum_{j=1}^n a_{ij}=1\) for every \(i=1,\dots, n\). A typical result (Theorem 3): if \(A\) is an \(n\times n\) stochastic primitive matrix with \(\exp(A)\geq (4n-3)n/5\), then \(A\) has a complex conjugate pair of eigenvalues with moduli exceeding \((1/2)^{1/s}\), where \(s=[\exp(A)-n]/ (n-2)\).