A direct proof for the matrix decomposition of chordal-structured positive semidefinite matrices (Q979016)

The paper deals with the matrix decomposition of chordal-structured positive semidefinite matrices. An undirected graph \(G\) is chordal if every cycle of length at least four has a chord. \textit{J. Agler, W. Helton, S. McCullough} and \textit{L. Rodman} [Linear Algebra Appl. 107, 101--149 (1988; Zbl 0655.15016)] proved that a graph \(G\) is chordal if and only if any positive semidefinite symmetric matrix, whose nonzero entries are specified by \(G\), can be decomposed as a sum of positive semidefinite matrices corresponding to the maximal cliques of \(G\). In this paper, the author shows an alternative proof of this result using simple linear algebra, which allows to impose an additional rank condition on positive semidefinite matrices obtained by the decomposition. In addition, the author presents a new proof, via duality theory, of the well-known characterization for positive semidefinite matrix completion of a chordal-structured matrix due to \textit{R. Grone, C. R. Johnson, E. M. Sá} and \textit{H. Wolkowicz} [Linear Algebra Appl. 58, 109--124 (1984; Zbl 0547.15011)].