arXiv++ Combinatorics

In the late 80s, V.~Arnold and V.~Vassiliev initiated the topological study of the space of real univariate polynomials of a given degree d and with no real roots of multiplicity exceeding a given positive integer. Expanding their studies, we consider the spaces of real monic univariate polynomials of degree d whose real divisors avoid sequences of root multiplicities taken from a given poset of compositions which is closed under certain natural combinatorial operations. In this paper, we concentrate on the fundamental group of such spaces. We find explicit presentations for the fundamental groups in terms of generators and relations and show that in a number of cases they are free with rank bounded from above by a quadratic function in d. We also show that the fundamental group stabilizes for d large. We further show that the fundamental groups admit an interpretation as special bordisms of immersions of 1-manifolds into the cylinder S^1 \times R, whose images avoid the tangency patterns from the poset with respect to the generators of the cylinder.

The classical problem of counting the number of real zeros of a real polynomial was solved a long time ago by Sturm. The analogous problem of counting the number of zeros that a polynomial has on the unit circle is, however, still an open problem. In this paper, we show that the second problem can be reduced to the first one through the use of a suitable pair of Möbius transformations - often called Cayley transformations - that have the property of mapping the unit circle onto the real line and vice versa. Although the method applies to arbitrary complex polynomials, we discuss in detail several classes of polynomials with symmetric zeros as, for instance, the cases of self-conjugate, self-adjoint, self-inversive, self-reciprocal and skew-reciprocal polynomials. Finally, an application of this method to Salem polynomials and to polynomials with small Mahler measure is also discussed.

We prove the following statement. Let $f\in\mathbb{R}[x_1,\ldots,x_d]$, for some $d\ge 3$, and assume that $f$ depends non-trivially in each of $x_1,\ldots,x_d$. Then one of the following holds. (i) For every finite sets $A_1,\ldots,A_d\subset \mathbb{R}$, each of size $n$, we have $$|f(A_1\times\ldots\times A_d)|=Ω(n^{3/2}), $$ with constant of proportionality that depends on ${\rm deg} f$. (ii) $f$ is of one of the forms \begin{align*} f(x_1,\ldots, x_d)&=h(p_1(x_1)+\cdots+p_d(x_d))~~\text{or}\\ f(x_1,\ldots, x_d)&=h(p_1(x_1)\cdot\ldots\cdot p_d(x_d)), \end{align*} for some univariate real polynomials $h(x)$, $p_1(x),\ldots,p_d(x)$. This generalizes the results from [ER00,RSS, RSdZ], which treat the cases $d=2$ and $d=3$.

A square matrix $M$ with real entries is said to be algebraically positive (AP) if there exists a real polynomial $p$ such that all entries of the matrix $p(M)>0$. A square sign pattern matrix $S$ is said to allow algebraic positivity if there is an algebraically positive matrix $M$ whose sign pattern class is $S$. On the other hand, $S$ is said to require algebraic positivity if any matrix $M$, having sign pattern class $S$, is algebraically positive. Motivated by open problems raised in the work of Kirkland, Qiao and Zhan (2016) on AP matrices, we list down all nonequivalent irreducible $3\times 3$ sign pattern matrices and classify each of them into three groups (i) those that require AP, (ii) those that allow but not require AP, or (iii) those that do not allow AP. We also give a necessary condition for an irreducible $n\times n$ sign pattern to allow algebraic positivity.

An $n\times n$ sign pattern $S$, which is a matrix with entries $0,+,-$, is called spectrally arbitrary if any monic real polynomial of degree $n$ can be realized as a characteristic polynomial of a matrix obtained by replacing the non-zero elements of $S$ by numbers of the corresponding signs. A sign pattern $S$ is said to be a superpattern of those matrices that can be obtained from $S$ by replacing some of the non-zero entries by zeros. We develop a new technique that allows us to prove spectral arbitrariness of sign patterns for which the previously known "Nilpotent Jacobian" method does not work. Our approach leads us to solutions of numerous open problems known in the literature. In particular, we provide an example of a sign pattern $S$ and its superpattern $S'$ such that $S$ is spectrally arbitrary but $S'$ is not, disproving a conjecture proposed in 2000 by Drew, Johnson, Olesky, and van den Driessche.

A great variety of fundamental optimization and counting problems arising in computer science, mathematics and physics can be reduced to one of the following computational tasks involving polynomials and set systems: given an $m$-variate real polynomial $g$ and a family of subsets $B$ of $[m]$, (1) find $S\in B$ such that the monomial in $g$ corresponding to $S$ has the largest coefficient in $g$, or (2) compute the sum of coefficients of monomials in $g$ corresponding to all the sets in $B$. Special cases of these problems, such as computing permanents, sampling from DPPs and maximizing subdeterminants have been topics of recent interest in theoretical computer science. In this paper we present a general convex programming framework geared to solve both of these problems. We show that roughly, when $g$ is a real stable polynomial with non-negative coefficients and $B$ is a matroid, the integrality gap of our relaxation is finite and depends only on $m$ (and not on the coefficients of g). Prior to our work, such results were known only in sporadic cases that relied on the structure of $g$ and $B$; it was not even clear if one could formulate a convex relaxation that has a finite integrality gap beyond these special cases. Two notable examples are a result by Gurvits on the van der Waerden conjecture for real stable $g$ when $B$ is a single element and a result by Nikolov and Singh for multilinear real stable polynomials when $B$ is a partition matroid. Our work, which encapsulates most interesting cases of $g$ and $B$, benefits from both - we were inspired by the latter in deriving the right convex programming relaxation and the former in establishing the integrality gap. However, proving our results requires significant extensions of both; in that process we come up with new notions and connections between stable polynomials and matroids which should be of independent interest.

We consider the problem of defining and computing real analogs of polynomial Hurwitz numbers, in other words, the problem of counting properly normalized real polynomials with fixed ramification profiles over real branch points. We show that, provided the polynomials are counted with an appropriate sign, their number does not depend on the order of the branch points on the real line. We study generating series for the invariants thus obtained, determine necessary and sufficient conditions for the vanishing and nonvanishing of these generating series, and obtain a logarithmic asymptotic for the invariants as the degree of the polynomials tends to infinity.

Let $F\in\mathbb{C}[x,y,s,t]$ be an irreducible constant-degree polynomial, and let $A,B,C,D\subset\mathbb{C}$ be finite sets of size $n$. We show that $F$ vanishes on at most $O(n^{8/3})$ points of the Cartesian product $A\times B\times C\times D$, unless $F$ has a special group-related form. A similar statement holds for $A,B,C,D$ of unequal sizes. This is a four-dimensional extension of our recent improved analysis of the original Elekes-Szabó theorem in three dimensions. We give three applications: an expansion bound for three-variable real polynomials that do not have a special form, a bound on the number of coplanar quadruples on a space curve that is neither planar nor quartic, and a bound on the number of four-point circles on a plane curve that has degree at least five.

Let $f(q)=a_rq^r+\cdots+a_sq^s$, with $a_r\neq 0$ and $a_s\neq 0$, be a real polynomial. It is a palindromic polynomial of darga $n$ if $r+s=n$ and $a_{r+i}=a_{s-i}$ for all $i$. Polynomials of darga $n$ form a linear subspace $\mathcal{P}_n(q)$ of $\mathbb{R}(q)_{n+1}$ of dimension $\lfloor{n/2}\rfloor+1$. We give transition matrices between two bases $\left\{q^j(1+q+\cdots+q^{n-2j})\right\}, \left\{q^j(1+q)^{n-2j}\right\}$ and the standard basis $\left\{q^j(1+q^{n-2j})\right\}$ of $\mathcal{P}_n(q)$. We present some characterizations and sufficient conditions for palindromic polynomials that can be expressed in terms of these two bases with nonnegative coefficients. We also point out the link between such polynomials and rank-generating functions of posets.

Given convex polytopes $P_1,...,P_r$ in $R^n$ and finite subsets $W_I$ of the Minkowsky sums $P_I=\sum_{i \in I} P_i$, we consider the quantity $N(W)=\sum_{I \subset {\bf [}r {\bf ]}} {(-1)}^{r-|I|} \big| W_I \big|$. We develop a technique that we call irrational mixed decomposition which allows us to estimate $N(W)$ under some assumptions on the family $W=(W_I)$. In particular, we are able to show the nonnegativity of $N(W)$ in some important cases. The quantity $N(W)$ associated with the family defined by $W_I=\sum_{i \in I} W_i$ is called discrete mixed volume of $W_1,...,W_r$. We show that for $r=n$ the discrete mixed volume provides an upper bound for the number of nondegenerate solutions of a tropical polynomial system with supports $W_1,...,W_n$. We also prove that the discrete mixed volume of $W_1,...,W_r$ is bounded from above by the Kouchnirenko number $\prod_{i=1}^r (|W_i|-1)$. For $r=n$ this number was proposed as a bound for the number of nondegenerate positive solutions of any real polynomial system with supports $W_1,...,W_n$. This conjecture was disproved, but our result shows that the Kouchnirenko number is a sharp bound for the number of nondegenerate positive solutions of real polynomial systems constructed by means of the combinatorial patchworking.

We completely characterize sections of the cones of nonnegative polynomials, convex polynomials and sums of squares with polynomials supported on circuits, a genuine class of sparse polynomials. In particular, nonnegativity is characterized by an invariant, which can be immediately derived from the initial polynomial. Furthermore, nonnegativity of such polynomials $f$ coincides with solidness of the amoeba of $f$, i.e., the Log-absolute-value image of the algebraic variety $\mathcal{V}(f) \subset (\mathbb{C}^*)^n$ of $f$. These results generalize earlier works both in amoeba theory and real algebraic geometry by Fidalgo, Kovacec, Reznick, Theobald and de Wolff and solve an open problem by Reznick. They establish the first direct connection between amoeba theory and nonnegativity of real polynomials. Additionally, these statements yield a completely new class of nonnegativity certificates independent from sums of squares certificates.

In this paper we characterize real bivariate polynomials which have a small range over large Cartesian products. We show that for every constant-degree bivariate real polynomial $f$, either $|f(A,B)|=Ω(n^{4/3})$, for every pair of finite sets $A,B\subset{\mathbb R}$, with $|A|=|B|=n$ (where the constant of proportionality depends on ${\rm deg} f$), or else $f$ must be of one of the special forms $f(u,v)=h(\varphi(u)+ψ(v))$, or $f(u,v)=h(\varphi(u)\cdotψ(v))$, for some univariate polynomials $\varphi,ψ,h$ over ${\mathbb R}$. This significantly improves a result of Elekes and Rónyai (2000). Our results are cast in a more general form, in which we give an upper bound for the number of zeros of $z=f(x,y)$ on a triple Cartesian product $A\times B\times C$, when the sizes $|A|$, $|B|$, $|C|$ need not be the same; the upper bound is $O(n^{11/6})$ when $|A|=|B|=|C|=n$, where the constant of proportionality depends on ${\rm deg} f$, unless $f$ has one of the aforementioned special forms. This result provides a unified tool for improving bounds in various Erd\H os-type problems in geometry and additive combinatorics. Several applications of our results to problems of these kinds are presented. For example, we show that the number of distinct distances between $n$ points lying on a constant-degree parametric algebraic curve which does not contain a line, in any dimension, is $Ω(n^{4/3})$, extending the result of Pach and de Zeeuw (2013) and improving the bound of Charalambides (2012), for the special case where the curve under consideration has a polynomial parameterization. We also derive improved lower bounds for several variants of the sum-product problem in additive combinatorics.

Hyperbolic polynomials are real polynomials whose real hypersurfaces are nested ovaloids, the inner most of which is convex. These polynomials appear in many areas of mathematics, including optimization, combinatorics and differential equations. Here we investigate the special connection between a hyperbolic polynomial and the set of polynomials that interlace it. This set of interlacers is a convex cone, which we write as a linear slice of the cone of nonnegative polynomials. In particular, this allows us to realize any hyperbolicity cone as a slice of the cone of nonnegative polynomials. Using a sums of squares relaxation, we then approximate a hyperbolicity cone by the projection of a spectrahedron. A multiaffine example coming from the Vamos matroid shows that this relaxation is not always exact. Using this theory, we characterize the real stable multiaffine polynomials that have a definite determinantal representation and construct one when it exists.

Let $\mathfrak{L}$ be a collection of $L$ lines in $\R^3$ and $J$ the set of joints formed by $\mathfrak{L}$, i.e. the set of points each of which lies in at least 3 non-coplanar lines of $\mathfrak{L}$. It is known that $|J| \lesssim L^{3/2}$ (first proved by Guth and Katz). For each joint $x \in J$, let the multiplicity $N(x)$ of $x$ be the number of triples of non-coplanar lines through $x$. We prove here that $\sum_{x \in J}N(x)^{1/2} \lesssim L^{3/2}$, while in the last section we extend this result to real algebraic curves of uniformly bounded degree in $\R^3$, as well as to curves in $\R^3$ parametrised by real polynomials of uniformly bounded degree.

The Wronski map is a finite, PGL_2(C)-equivariant morphism from the Grassmannian Gr(d,n) to a projective space (the projectivization of a vector space of polynomials). We consider the following problem. If C_r < PGL_2(C) is a cyclic subgroup of order r, how may C_r-fixed points are in the in a fibre of the Wronski map over a C_r-fixed point in the base? In this paper, we compute a general answer in terms of r-ribbon tableaux. When r=2, this computation gives the number of real points in the fibre of the Wronski map over a real polynomial with purely imaginary roots. More generally, we can compute the number of real points in certain intersections of Schubert varieties. When r divides d(n-d) our main result says that the generic number of C_r-fixed points in the fibre is the number of standard r-ribbon tableaux rectangular shape (n-d)^d. Computing by a different method, we show that the answer in this case is also given by the number of of standard Young tableaux of shape (n-d)^d that are invariant under N/r iterations of jeu de taquin promotion. Together, these two results give a new proof of Rhoades' cyclic sieving theorem for promotion on rectangular tableaux. We prove analogous results for dihedral group actions.

We explore the consequences of an ideal I of real polynomials having a real radical initial ideal, both for the geometry of the real variety of I and as an application to sums of squares representations of polynomials. We show that if in_w(I) is real radical for a vector w in the tropical variety, then w is in the logarithmic set of the real variety. We also give algebraic sufficient conditions for w to be in the logarithmic limit set of a more general semialgebraic set. If in addition the entries of w are positive, then the corresponding quadratic module is stable. In particular, if in_w(I) is real radical for some positive vector w then the set of sums of squares modulo I is stable. This provides a method for checking the conditions for stability given by Powers and Scheiderer.

Let $P_N(R)$ be the space of all real polynomials in $N$ variables with the usual inner product $<, >$ on it, given by integrating over the unit sphere. We start by deriving an explicit combinatorial formula for the bilinear form representing this inner product on the space of coefficient vectors of all polynomials in $P_N(R)$ of degree $\leq M$. We exhibit two applications of this formula. First, given a finite dimensional subspace $V$ of $P_N(R)$ defined over $Q$, we prove the existence of an orthogonal basis for $(V, <, >)$, consisting of polynomials of small height with integer coefficients, providing an explicit bound on the height; this can be viewed as a version of Siegel's lemma for real polynomial inner product spaces. Secondly, we derive a criterion for a finite set of points on the unit sphere in $R^N$ to be a spherical $M$-design.

Let $L$ be a finite distributive lattice and $μ: L \to {\mathbb R}^{+}$ a log-supermodular function. For functions $k: L \to {\mathbb R}^{+}$ let $$E_μ (k; q) \defeq \sum_{x\in L} k(x) μ(x) q^{{\mathrm rank}(x)} \in {\mathbb R}^{+}[q].$$ We prove for any pair $g,h: L\to {\mathbb R}^{+}$ of monotonely increasing functions, that $$E_μ (g; q)\cdot E_μ (h; q) \ll E_μ (1; q)\cdot E_μ (gh; q), $$ where ``$ \ll $'' denotes coefficientwise inequality of real polynomials. The FKG inequality of Fortuin, Kasteleyn and Ginibre (1971) is the real number inequality obtained by specializing to $q=1$. The polynomial FKG inequality has applications to $f$-vectors of joins and intersections of simplicial complexes, to Betti numbers of intersections of certain Schubert varieties, and to the following kind of correlation inequality for power series weighted by Young tableaux. Let $Y$ be the set of all integer partitions. Given functions $k, μ: Y \rarr \R^+$, and parameters $0\le s\le t$, define the formal power series $$F_μ(k ; z) \defeq \sum_{\la\in Y} k(\la) μ(\la) (f_{\la})^t \frac{z^{|\la|}}{(|\la| !)^s} \in \R^+ [[z]], $$ %\sum_{\la\in Y} k(\la) μ(\la) (f_{\la})^t \frac{z^{|\la|}}{|\la| !} \in \R^+ [[z]],$$ where $f_{\la}$ is the number of standard Young tableaux of shape $\la$. Assume that $μ: Y\rarr \R^+$ is log-supermodular, and that $g, h: Y \rarr \R^+$ are monotonely increasing with respect to containment order of partition shapes. Then $$F_μ(g;z) \cdot F_μ(h;z) \ll F_μ(1;z) \cdot F_μ(gh;z). $$

We investigate the Cesaro and Abel sums of divergent series of the form $\sum_{n\geq 0} a_n T^nP(x)$, where $P$ is a real polynomial and $T$ is a translation invariant operator on the space of real polynomials.

We describe all linear operators on spaces of multivariate polynomials preserving the property of being non-vanishing in open circular domains. This completes the multivariate generalization of the classification program initiated by Pólya-Schur for univariate real polynomials and provides a natural framework for dealing in a uniform way with Lee-Yang type problems in statistical mechanics, combinatorics, and geometric function theory. This is an announcement with some of the main results in arXiv:0809.0401 and arXiv:0809.3087.