Linear bijective maps preserving fixed products of matrices (Q1984355): Difference between revisions

From MaRDI portal
Set OpenAlex properties.
ReferenceBot (talk | contribs)
Changed an Item
 
(One intermediate revision by one other user not shown)
Property / Wikidata QID
 
Property / Wikidata QID: Q123296246 / rank
 
Normal rank
Property / cites work
 
Property / cites work: On maps characterized by action on equal products / rank
 
Normal rank
Property / cites work
 
Property / cites work: On maps preserving products of matrices / rank
 
Normal rank
Property / cites work
 
Property / cites work: On maps preserving products equal to fixed elements / rank
 
Normal rank
Property / cites work
 
Property / cites work: Linear Transformations on Algebras of Matrices: The Invariance of the Elementary Symmetric Functions / rank
 
Normal rank

Latest revision as of 15:52, 26 July 2024

scientific article
Language Label Description Also known as
English
Linear bijective maps preserving fixed products of matrices
scientific article

    Statements

    Linear bijective maps preserving fixed products of matrices (English)
    0 references
    0 references
    16 September 2021
    0 references
    Let \(\mathcal{M}=\mathcal{M}_{n}\) be the algebra of all \(n\times n\) complex matrices, and fix \(M,N\in\mathcal{M}\). In this paper, the linear bijective maps \(\phi:\mathcal{M}\to\mathcal{M}\) such that \(\phi(A)\phi(B)=M\) whenever \(AB=N\) are described explicitly. Namely, when \(M\) or \(N\) is not invertible, such a \(\phi\) is given exactly by \(\phi=\phi_{\alpha,U}\) for some some nonzero complex number \(\alpha\) and some invertible matrix \(U\) in \(\mathcal{M}\) with \(\alpha^{2}UN=MU\), where \(\phi_{\alpha,U}:\mathcal{M}\to\mathcal{M}\) is defined by \(\phi_{\alpha,U}(T)=\alpha UTU^{-1}\) for all \(T\in\mathcal{M}\); when \(M\) and \(N\) are both invertible, then such a \(\phi\) is precisely a \(\phi_{\alpha,U}\) or a \(\psi_{\beta,V}\), where \(\beta\in\mathbb{C}\) is nonzero, \(V\in\mathcal{M}\) invertible with \(\beta^{2}VN^{\text{t}}=M^{-1}V\), and \(\psi_{\beta,V}:\mathcal{M}\to\mathcal{M}\) is defined by \(\psi_{\beta,V}(T)=\beta VT^{\text{t}}V^{-1}\) for all \(T\in\mathcal{M}\). In order to get the explicit form of particular linear maps on \(\mathcal{M}\), the following matrix equations are solved. Let \(C\in\mathcal{M}\) and let \(\mathcal{P}=\{(A,B)\in\mathcal{M}^{2}\mid AB=C\}\). (1) If \(X\in\mathcal{M}\) is such that \(AXB=XC\) for all \((A,B)\in\mathcal{P}\), then \(X=\alpha I_{n}\) for some \(\alpha\in\mathbb{C}\); (2) if \(Y\in\mathcal{M}_{n}\) is such that \(BYA=CY\) for all\((A,B)\in\mathcal{P}\), then \(YC=CY=\beta I_{n}\) for some \(\beta\in\mathbb{C}\).
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    linear preserver problem
    0 references
    products of matrices
    0 references
    0 references
    0 references