Total positivity and Neville elimination (Q1184469): Difference between revisions

Let \(A\) be an \(n\times m\) matrix, then according to the authors, Neville elimination consists of the reduction of \(A\) to echolon form by producing zeros in successive columns of \(A\) by adding to each row an appropriate multiple of the previous row. (Some re-orderings of the rows may be needed to make it possible.) This method of elimination is described in terms of Schur complements. It is used to prove a sign characterization of the total positivity of non-singular matrices by means of the signs of some of the minors. (A matrix is (strictly) totally positive if all the minors of all orders are (positive) non-negative.) There is a characterization of strictly positive matrices due to \textit{M. Fekete} and \textit{G. Pólya} [Über ein Problem von Laguerre. Rend. Circ. Mat. Palermo 34, 89-120 (1912; JFM 43.0145.01)] which requires the testing of all minors from consecutive rows and columns. It is shown here that the labour of checking can be considerably reduced to that of checking signs of minors from consecutive rows and columns which include the first row or column. The paper ends with a characterization of arbitrary totally positive matrices in terms of Neville elimination. The authors give other characterizations of totally positive non-singular matrices.