A CMV-BASED EIGENSOLVER FOR COMPANION MATRICES

作者：Bevilacqua R^*; Del Corso G M; Gemignani L

来源：SIAM Journal on Matrix Analysis and Applications, 2015, 36(3): 1046-1068.

DOI：10.1137/140978065

• companion matrix
• QR eigenvalue algorithm
• CMV matrix
• rank structure
• complexity
• accuracy

In this paper we present a novel matrix method for polynomial rootfinding. We approximate the roots by computing the eigenvalues of a permuted version of the companion matrix associated with the polynomial. This form, referred to as a lower staircase form of the companion matrix in the literature, has a block upper Hessenberg shape with possibly nonsquare subdiagonal blocks. It is shown that this form is well suited to the application of the QR eigenvalue algorithm. In particular, each matrix generated under this iteration is block upper Hessenberg and, moreover, all its submatrices located in a specified upper triangular portion are of rank two at most, with entries represented by means of four given vectors. By exploiting these properties we design a fast and computationally simple structured QR iteration which computes the eigenvalues of a companion matrix of size n in lower staircase form using O(n(2)) flops and O(n) memory storage. So far, this iteration is theoretically faster than the fastest variant of the QR iteration for companion matrices in customary Hessenberg form. Numerical experiments show efficiency and accuracy of the proposed approach.

• 出版日期2015