Supersolvable frame-matroid and graphic-lift lattices (Q1840839): Difference between revisions

The author investigates sufficient conditions (supersolvability) for a complete integral factorization of the characteristic polynomials of geometric lattices that correspond to arrangements of hyperplanes and finite matroids. Recall that a geometric lattice of rank \(n\) is supersolvable if there exists a saturated chain \( 0 = x_{0} < x_{1} < \cdots < x_{n-1} < x_{n} = 1\) with rank \((x_{i})= i\) such that every flat \(x_{i}\) is modular. This interesting paper includes new results on the supersolvability of lattices. In part it is a continuation of the article [\textit{R. P. Stanley}, Algebra Univers. 2, 197-217 (1972; Zbl 0256.06002)]. The paper under review also includes a concisely-written survey of recent results on biased and gain graphs, biased and lift matroids, coloring and polynomials (Section 1). The main results of the paper are theorems about supersolvability of two types of geometric lattices: (i) lattices whose matroid is a frame; (ii) lattices defined by a graphic lift matroid. The author also gives the geometric interpretation of his theorems in terms of arrangements in \(F^n\), where \(F\) is a skew field. Arrangements of hyperplanes connect the integral representation of hypergeometric functions and Orlik-Solomon algebras with conformal field theory, representation theory, algebraic \(K\)-theory and algebraic geometry. A clear and informative overview of these connections can be found in lectures notes by \textit{A. Varchenko} [Proc. Int. Congr. Math., Kyoto/Japan 1990, Vol.~I, 281-300 (1991; Zbl 0747.33002)] and \textit{P. Orlik} and \textit{H. Terao} [Arrangements and hypergeometric functions (MSJ Memoirs 9, Tokio: Mathematical Society of Japan) (2001)]. For matroid theory the author follows \textit{J. G. Oxley} [Matroid theory (Oxford Science Publications. Oxford: Oxford University Press) (1992; Zbl 0784.05002)]. The history of matroid theory is described in the anthology [\textit{J. P. S. Kung}, A source book in matroid theory (Boston, Basel, Stuttgart: Birkhäuser) (1986; Zbl 0597.05019)]. The organization of the paper under review is as follows: Introduction. 1. Biased graphs, matroids, etc. 2. Frames. 3. Graphic lifts. 4. Examples. 5. Comments and questions. The author calls the matroid of a geometric lattice (of finite rank) such that it has a basis, or can be extended to have one, and such that each point lies on a line generated by a pair of basis elements a (finite rank) frame matroid. The bias matroid \(G(\Omega)\) is the matroid on the set of edges of a biased graph \(\Omega\) whose circuits are the bias circuits. Theorem 2.2. Let \(\Omega\) be a simply biased graph of finite order. \(G(\Omega)\) is supersolvable if and only if each connected component of \(\Omega\) either: (i) has a bias-simplicial vertex ordering; or (ii) is a simplicial extension of one of (a) \((mK_{2},0),\) when \(m \geq 2,\) or \(\langle \pm K_{3} \rangle,\) or \(\langle \Sigma \rangle\) for \(\Sigma = +\Gamma \cup -S_{k},\) where \(\Gamma\) is a chordal simple graph of finite order, \(S_{k}\) is a \(k\)-edge star whose vertex set lies in \(V(\Gamma),\) and the noncentral vertices of \(S_{k}\) are a clique in \(\Gamma\). Furthermore, the bias-simplicial vertex ordering (or simplicial extension) can be chosen so that any desired bias-simplicial (or simplicial) vertex is the last vertex. Section 3 contains the main theorems. For a biased graph \(\Omega\) the author defines the extended lift matroid \(L_{0}(\Omega)\) and the lift matroid \(L(\Omega).\) Theorem 3.2. Let \(\Omega\) be a simply biased, connected link graph of finite order. (A) \(L_{0}(\Omega)\) is supersolvable if and only if \(\Omega\) has a link-simplicial vertex ordering. (B) \(L(\Omega)\) is supersolvable if and only if: (i) \(\Omega\) is balanced and \(\parallel \Omega \parallel\) is chordal; or (ii) \(\Omega\) is as in Theorem 2.2 (ii). Furthermore, the link-simplicial vertex ordering (or simplicial extension) can be chosen so that any desired bias-simplicial (or simplicial) vertex is last. The author applies his results to three kinds of examples and to an extension of a theorem of \textit{P. H. Edelman} and \textit{V. Reiner} [Math. Z. 215, No.~3, 347-365 (1994; Zbl 0793.05122)].