Kostka functions associated to complex reflection groups (Q514192): Difference between revisions

The Kostka functions \(K_{\lambda,\mu}(t)\in\mathbb{Z}[t]\), indexed by pairs of partitions \(\lambda,\mu\vdash n\), are defined as the entries in the transition matrix from Schur functions to Hall-Littlewood functions in the ring of symmetric functions with one parameter \(\Lambda^n[t]\); for further discussion see Chapter III, Section 6 of \textit{I.G. Macdonald}'s book [Symmetric functions and Hall polynomials. Oxford: Oxford University Press (2015; Zbl 1332.05002)]. A geometric interpretation of the Kostka functions was given by \textit{G. Lusztig} [Adv. Math. 42, 169--178 (1981; Zbl 0473.20029)] who proved that a certain modification of the Kostka function \(K_{\lambda,\mu}(t)\) (i.e. a Green function) is the Poincaré polynomial of the stalk of a certain sheaf on a certain algebraic variety \(\overline{\mathcal{O}}_{\lambda}\) (i.e. the closure of the conjugacy class of unipotent operators on a vector space \(V\) with Jordan block decomposition corresponding to \(\lambda\)) at a point \(x\in\mathcal{O}_{\mu}\subset\overline{\mathcal{O}}_{\lambda}\). In particular, Lusztig's theorem implies that the \(\mathbb{Z}\)-coefficients of the Kostka functions \(K_{\lambda,\mu}(t)\) are non-negative. Another, more combinatorial, way to see that the coefficients of \(K_{\lambda,\mu}(t)\) are non-negative is by the Lascoux-Schützenberger formula [\textit{A. Lascoux} and \textit{M.P. Schützenberger}, Acad. Sci. Paris 286A, 323--324 (1978; Zbl 0374.20010)]: \[ K_{\lambda,\mu}(t)=\sum_{T}t^{c(T)} \] where the sum runs over all tableaux \(T\) of shape \(\lambda\) and weight \(\mu\), and the \(c(T)\) is a combinatorial invariant called the charge of \(T\). In a previous paper [\textit{T. Shoji}, J. Algebra 245, 650--694 (2001; Zbl 0997.20044)], the author of the paper under review has defined analogues of Schur functions, Hall-Littlewood functions, and Kostka functions associated to the complex reflection group \(G(r,1,n)=\mathfrak{S}_n\rtimes\left(\mathbb{Z}/r\mathbb{Z}\right)^n\). In the paper under review, the author shows that his ``generalized'' Kostka functions have a geometric interpretation, proving an analogue of Lusztig's theorem (Theorem 2.6), as well as a combinatorial interpretation by proving an analogue of the Lascoux-Shützenberger formula (Theorem 3.14). The author's replacement for Lusztig's orbit decomposition of the unipotent part of \(\mathrm{GL}(V)\) is a decomposition of the unipotent part of something he calls the \textit{enhanced variety of level \(r\)} which may be of independent interest. As the author points out, the \(r=2\) case for these results was previously obtained by \textit{P.N. Achar} and \textit{A. Henderson} [Adv. Math. 219 (1), 27--62 (2008; Zbl 1205.14061)].