On weakly positive unit forms (Q1113972)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On weakly positive unit forms |
scientific article |
Statements
On weakly positive unit forms (English)
0 references
1988
0 references
An integral quadratic form q: \({\mathbb{Z}}^ n\to {\mathbb{Z}}\) given by \(q(x)=\sum^{n}_{a=1}x^ 2_{a}+\sum_{a<b}q_{ab}x_ ax_ b\) is called a unit form. In a natural way, with every unit form is associated a nonoriented multigraph with the edges of two kinds and no loops. q is said to be weakly positive if \(q(x)>0\) whenever \(x\neq 0\) and \(x_{a}\geq 0\) for all a; q is said to be a critical form if it is not weakly positive but every form obtained from q by fixing some of the \(x_ a's\) equal to zero is weakly positive. The author obtains a criterion for weak positivity of unit forms by giving a combinatorial classification of the critical forms in terms of the associated multigraphs. The results of the paper generalize those of \textit{S. A. Ovsienko} [Integral weakly positive forms, in Schurian matrix problems and quadratic forms, Prepr. Inst. Math. Acad. Sci. Ukrainian SSR, Kiev, (Russian), 3-17 (1978)]. In addition, the author presents another approach to the results of \textit{K. Bongartz}, [Manuscr. Math. 46, 117-136 (1984; Zbl 0537.16024)] and \textit{D. Happel} and \textit{D. Vossieck} [ibid. 42, 221-243 (1983; Zbl 0516.16023)]. A criterion for weak positivity of an arbitrary real quadratic form p: \({\mathbb{R}}^ n\to {\mathbb{R}}\), given by \(p(x)=\sum^{n}_{a,b=1}p_{ab}x_ ax_ b\), where \(p_{ab}=p_{ba}\) for all a, b, was obtained by \textit{M. V. Zel'dich} [Linear algebra and representation theory, Collect. Sci. Works, Kiev 1983, 135-137 (1983; Zbl 0583.15009)]. Since the latter paper is not easily accessible to the English-speaking reader, it seems worthwhile to state its main result. Namely, p is weakly positive if and only if every symmetric submatrix S of its matrix \(P=[p_{ab}]\), including P itself, satisfies either of the following two conditions: (i) \(\det (S)>0\), (ii) there exists an element of S whose cofactor (in S) is nonpositive. It might be interesting to investigate whether the above statement can be used to simplify the proofs in the paper under review.
0 references
quadratic form
0 references
unit form
0 references
nonoriented multigraph
0 references
critical form
0 references
weak positivity
0 references
combinatorial classification
0 references