The Smith normal form (Q677923): Difference between revisions

The author shows that the Smith normal form can be treated in rather general terms of matrix equivalence in the following manner. Let \({\mathfrak R}\) be a commutative ring with identity 1 and \(A, B\in{\mathfrak R}_{m,n}\) (ring of \(m\times n\) matrices over \({\mathfrak R}\)). \(A\sim B\) if \(U\), \(V\) exist such that \(U\in\text{GL}(m,{\mathfrak R})\), \(V\in\text{GL}(n,{\mathfrak R})\) and \(B=UAV\). For the purpose of the paper one considers the cases: \({\mathfrak R}=\mathbb{Z}\), \({\mathfrak R}=\mathbb{F}\) (a field) and \({\mathfrak R}=\mathbb{F}[x]\) (the ring of polynomials) and \(m=n\) for which the following problems are addressed: Smith theorem, determinantal divisors, elementary divisors, divisibility properties, and applications to the solutions of linear systems, permutation equivalence, and Abelian group theory. Possible generalizations are finally examined.