Polytopes with Mass Linear Functions, Part I
نویسندگان
چکیده
We analyze mass linear functions on simple polytopes ∆, where a mass linear function is an affine function on ∆ whose value on the center of mass depends linearly on the positions of the supporting hyperplanes. On the one hand, we show that certain types of symmetries of ∆ give rise to nonconstant mass linear functions on ∆. We call mass linear functions which arise in this way inessential; the others are called essential. On the other hand, we show that most polytopes do not admit any nonconstant mass linear functions. Further, if every affine function is mass linear on ∆, then ∆ is affine equivalent to a product of simplices. Our main result is a classification of all 2-dimensional simple polytopes and 3-dimensional smooth polytopes which admit nonconstant mass linear functions. In particular there is only one family of smooth polytopes of dimension ≤ 3 which admit essential mass linear functions. In part II, we will complete this classification in the 4dimensional case. These results have geometric implications. Fix a symplectic toric manifold (M,ω, T,Φ) with moment polytope ∆ = Φ(M). Let Symp0(M,ω) denote the identity component of the group of symplectomorphisms of (M,ω). Any linear function H on ∆ generates a Hamiltonian R action onM whose closure is a subtorus TH of T . We show that if the map π1(TH) → π1(Symp0(M,ω)) has finite image, then H is mass linear. By the claims described above, this implies that in most cases the induced map π1(T )→ π1(Symp0(M,ω)) is an injection. Moreover, the map does not have finite image unless M is a product of projective spaces. Note also that there is a natural maximal compact connected subgroup Isom0(M) ⊂ Symp0(M,ω); there is a natural compatible complex structure J on M , and Isom0(M) is the identity component of the group of symplectomorphisms that also preserve this structure. We prove that if the polytope ∆ supports no nontrivial essential mass linear functions, then the induced map π1(Isom0(M)) → π1(Symp0(M,ω)) is injective. Therefore, this map is injective for all 4-dimensional symplectic toric manifolds. It is also injective in the 6-dimensional case unless M is a CP 2 bundle over CP . with Appendix written with V. Timorin
منابع مشابه
Linear Programming, the Simplex Algorithm and Simple Polytopes
In the first part of the paper we survey some far reaching applications of the basis facts of linear programming to the combinatorial theory of simple polytopes. In the second part we discuss some recent developments concurring the simplex algorithm. We describe sub-exponential randomized pivot roles and upper bounds on the diameter of graphs of polytopes.
متن کاملA Characteristic Number of Bundles Determined by Mass Linear Pairs
Abstract. Let ∆ be a Delzant polytope in Rn and b ∈ Zn. Let E denote the symplectic fibration over S determined by the pair (∆, b). We prove the equivalence between the fact that (∆, b) is a mass linear pair (D. McDuff, S. Tolman, Polytopes with mass linear functions, part I. arXiv:0807.0900 [math.SG]) and the vanishing of a characteristic number of E in the following cases: When ∆ is a ∆n−1 bu...
متن کاملSolving multiobjective linear programming problems using ball center of polytopes
Here, we aim to develop a new algorithm for solving a multiobjective linear programming problem. The algorithm is to obtain a solution which approximately meets the decision maker's preferences. It is proved that the proposed algorithm always converges to a weak efficient solution and at times converges to an efficient solution. Numerical examples and a simulation study are used to illu...
متن کاملExponential membership function and duality gaps for I-fuzzy linear programming problems
Fuzziness is ever presented in real life decision making problems. In this paper, we adapt the pessimistic approach tostudy a pair of linear primal-dual problem under intuitionistic fuzzy (I-fuzzy) environment and prove certain dualityresults. We generate the duality results using exponential membership and non-membership functions to represent thedecision maker’s satisfaction and dissatisfacti...
متن کاملOn the Links between Probabilistic Graphical Models and Submodular Optimisation. (Liens entre modèles graphiques probabilistes et optimisation sous-modulaire)
A probabilistic graphical model encodes conditional independences among random variables, which is related to factorisable distributions. Moreover, the entropy of a probability distribution on a set of discrete random variables is always bounded by the entropy of its factorisable counterpart. This is due to the submodularity of entropy on the set of discrete random variables. Submodular functio...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2008