Wronskians, total positivity, and real Schubert calculus (Q6161646): Difference between revisions

Let \(Fl_n (\mathbb R)\) denote the complete flag variety, consisting of all sequences \(V = (V_1 \subset \ldots \subset V_{n-1})\) of nested subspaces of \(\mathbb R^n\) such that each \(V_k\) has dimension \(k\). \textit{G. Lusztig} [Prog. Math. 123, 531--568 (1994; Zbl 0845.20034)] introduced two remarkable subsets of \(Fl_n (\mathbb R)\) , called the totally nonnegative part \({Fl_n (\mathbb R)}^{\geq 0}\) and the totally positive part \({Fl_n (\mathbb R)}^{>0}\). They may be defined as the subsets where all Plücker coordinates are nonnegative or positive, respectively, up to rescaling. The totally nonnegative parts of flag varieties have been widely studied, with connections to representation theory, combinatorics and cluster algebras, high-energy physics, topology, and many other topics. On the other hand, the Wronskian of a sequence of sufficiently differentiable functions \(f_1 (x), \ldots , f_k (x)\) (defined on \(\mathbb R\) or \(\mathbb C\)) can be defined easily. The Wronskian is identically zero if \(f_1, \ldots , f_k\) are linearly dependent; otherwise, it only depends on the linear span \(V\) of \(f_1, \ldots , f_k\) up to multiplication by a nonzero scalar. In particular, it makes sense to write \(W_r (V)\), and its zeros are well-defined. In this paper they introduce a new connection between total positivity and Wronskians, and use it to show the hidden role that total positivity plays in certain conjectures in real Schubert calculus. Also they present some conjectures in real Schubert calculus.