Nonnegative realization of spectra having negative real parts (Q2497247)

The nonnegative inverse eigenvalue problem is the problem of determining conditions for a list of numbers \(\sigma\) to be the spectrum of a nonnegative matrix. The authors give the following solution: Theorem: Let \(\lambda_2\), \(\lambda_3\), \dots , \(\lambda_n\) be nonzero complex numbers with real part less than or equal to zero and let \(\rho\) be a positive real number. Then the list \((\rho, \lambda_2, \lambda_3, \dots , \lambda_n)\) is the nonzero spectrum of a nonnegative matrix if the following assumptions are satisfied: (1) the list is closed under complex conjugation; (2) \(\rho + \sum_{i=2}^n \lambda_i \geq 0\); (3) \(\rho^2 + \sum_{i=2}^n \lambda_i^2 > 0\). Theorem: Let \(\lambda_2\), \(\lambda_3\), \dots , \(\lambda_n\) be nonzero complex numbers with real part less than or equal to zero and let \(\rho\) be a positive real number. Then the list \((\rho, \lambda_2, \lambda_3, \dots, \lambda_n)\) is the spectrum of a nonnegative matrix if and only if the following assumptions are satisfied: (1) the list is closed under complex conjugation; (2) \(s_1=\rho + \sum_{i=2}^n \lambda_i \geq 0\); (3) \(s_2=\rho^2 + \sum_{i=2}^n \lambda_i^2 > 0\); (4) \(s_1^2 \leq ns_2\).