Sign patterns occurring in orthogonal matrices
The problem of characterizing square sign patterns that allow orthogonality is considered. The ``obvious necessary condition -- that every pair of rows of the sign pattern must allow orthogonal vectors -- is also sufficient for sign patterns of dimension \(n=4\). One of the authors has previously shown the falsity of the conjecture that this condition is sufficient for any dimension by providing a counterexample for \(n=5\). This article gives two counterexamples for \(\pm\) sign patterns of dimension \(n=6\). The main result of the article consists in two new broad necessary conditions for the orthogonal sign pattern problem, which significantly generalize the obvious necessary condition. These new conditions explain why the given counterexamples do not allow orthogonality.
- Sparse orthogonal matrices.
- Sign patterns of \(J\)-orthogonal matrices
- Sign patterns that allow diagonalizability
- Sign pattern matrices that allow orthogonality
- Sign patterns of orthogonal matrices and the strong inner product property
- Sign-nonsingular matrices and orthogonal sign-patterns
- Essential sign change numbers of full sign pattern matrices
- The orthogonal signature
- Regular sign potentially orthogonal sign patterns
- Rational orthogonal versus real orthogonal
- ± sign pattern matrices that allow orthogonality
- Patterns of alternating sign matrices
- Symmetric sign pattern matrices that require unique inertia
- Orthogonal realizations of random sign patterns and other applications of the SIPP
- The number of negative entries in a sign pattern allowing orthogonality
- The strong singular value property for matrices
- The number of equivalence classes of symmetric sign patterns
This page was built for publication: Sign patterns occurring in orthogonal matrices
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4222257)