lattice simplices of maximal dimension with a given degree
play

Lattice simplices of maximal dimension with a given degree Akihiro - PowerPoint PPT Presentation

Lattice simplices of maximal dimension with a given degree Akihiro Higashitani (Kyoto Sangyo University) Einstein Workshop on Lattice Polytopes at Berlin 1216 Dec. 2016 1 Contents 1. Introduction to Cayley Conjecture and Nills bound 2.


  1. Lattice simplices of maximal dimension with a given degree Akihiro Higashitani (Kyoto Sangyo University) Einstein Workshop on Lattice Polytopes at Berlin 12–16 Dec. 2016 1

  2. Contents 1. Introduction to Cayley Conjecture and Nill’s bound 2. Correspondence between lattice simplices and finite abelian groups (Johannes’ talk) 3. The case d + 1 = 4 s − 1 and Cayley Conjecture 4. The case d + 1 = f (2 s ) and Cayley Conjecture (j.w.w. K. Kashiwabara) 2

  3. 1.1. Introduction to Cayley Conjecture Let P ⊂ R d be a lattice polytope , i.e., P is a convex polytope whose vertices are the points in Z d . P ◦ : the interior of P dim P = d • codeg( P ) := min { k : kP ◦ ∩ Z d � = ∅} • deg( P ) := d + 1 − codeg( P ) Example P (0,0,1) codeg( P ) = 3 deg( P ) = 3 + 1 − 3 = 1 (0,1,0) (1,0,0) (1,1,0) 3

  4. Why do we say deg( P ) degree of P ? Remark For a lattice polytope P ⊂ R d , we consider the n ≥ 0 | nP ∩ Z d | t n . Then this becomes a rational Ehrhart series � function of the form h ∗ P ( t ) | nP ∩ Z d | t n = � (1 − t ) d +1 , n ≥ 0 where h ∗ P ( t ) is a polynomial in t . We say that h ∗ P ( t ) is the h ∗ -polynomial of P . (the degree of h ∗ ( t )) = deg( P ) . 4

  5. For a lattice polytope P ⊂ R d , a lattice pyramid over P is defined by Pyr( P ) := conv( { ( α, 0) ∈ R d +1 : α ∈ P }∪{ (0 , . . . , 0 , 1) } ) ⊂ R d +1 . Then dim(Pyr( P )) = dim P + 1. In particular, those are not unimod. equiv., however... d+1 R Remark Pyr(P) We have h ∗ P ( t ) = h ∗ Pyr( P ) ( t ), in R d particular, deg( P ) = deg(Pyr( P )) ✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿ . P 5

  6. Motivation We want to know Cayley structure of lattice polytopes. ✓ ✏ Cayley Polytope • P 0 , P 1 , . . . , P ℓ ⊂ R d : lattice polytopes P 0 ∗ P 1 ∗· · ·∗ P ℓ := conv(( P 0 × 0 ) ∪ ( P 1 × e 1 ) ∪· · ·∪ ( P ℓ × e ℓ )) ⊂ R d + ℓ We say P 0 ∗ · · · ∗ P ℓ is a Cayley polytope . R d + ℓ , • For a lattice polytope P ⊂ when there exist P 0 , P 1 , . . . , P ℓ ⊂ R d s.t. P ∼ = P 0 ∗ · · · ∗ P ℓ , we say P 0 ∗ · · · ∗ P ℓ is a Cayley decomposition of P . • For a lattice polytope P , let C ( P ) := max( { ℓ +1 : ∃ P 0 , . . . , ∃ P ℓ s.t. P ∼ = P 0 ∗· · ·∗ P ℓ } ) . ✒ ✑ 6

  7. ✓ ✏ (Strong) Cayley Conjecture (Dickenstein–Nill ’12) Let P be a lattice polytope of dimension d with degree s . ⇒ C ( P ) ≥ d + 1 − 2 s . d > 2 s = ✒ ✑ ✓ ✏ (Weak) Cayley Conjecture Let P be a lattice polytope of dimension d with degree s . ⇒ C ( P ) ≥ 2, d > 2 s = namely, P can be just decomposed into at least two polytopes. ✒ ✑ Strong Cayley conjecture is true if • P : smooth (Dickenstein–Nill ’10) • P : Gorenstein (DiRocco–Haase–Nill–Paffenholz ’13) • some class of (0 , 1)-polytopes? (work in progress) 7

  8. Theorem (Haase–Nill–Payne ’09) Let P be a lattice polytope of dimension d with degree s . ⇒ C ( P ) ≥ d + 1 − ( s 2 + 19 s − 4) / 2 d > ( s 2 + 19 s − 4) / 2 = Remark ∃ counterexample (appear later) for strong Cayley Conjecture The existence of counterexample for weak Cayley Conjecture might be still open. − → I want to know C ( P ) in order to give its “sharp” bound. I expect the bound of C ( P ) can be given like d + 1 − (linear of s ) . 8

  9. 1.2. (modified) Nill’s bound On the other hand, the following theorem is known: ∞ � m � � For m ∈ Z > 0 , let f ( m ) = . 2 ℓ ℓ =0 Theorem (Nill 2008, H. 2016) P : lattice simplex of dimension d with degree s ⇒ d + 1 ≤ f (2 s ) ≤ 4 s − 1 P is NOT a lattice pyramid = Moreover, f (2 s ) is sharp but f (2 s ) < 4 s − 1 in general. (Explain later more precisely.) 9

  10. Thus, it is natural to study the following problem: ✓ ✏ Problem Give a complete characterization of lattice simplices of d + 1 = f (2 s ) . dimension d with degree s satisfying ✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿ ✒ ✑ Remark • A complete characterization of lattice polytopes of degree 1 which are not lattice pyramids was given by Batyrev–Nill (2007). simplices of degree 2 • A complete characterization of lattice ✿✿✿✿✿✿✿✿ which are not lattice pyramids was given by H.–Hofscheier (2016+). 10

  11. 2. Correspondence between lattice simplices and finite abelian groups ✓ ✏ We will review the correspondence between unimodular equvalence classes of lattice simplices and finite abelian groups . ✒ ✑ ∆ ⊂ R d : lattice simplex of dimension d v 0 , v 1 , . . . , v d ∈ Z d : vertices of ∆ � � d d ( x 0 , x 1 , . . . , x d ) ∈ ( R / Z ) d +1 : x i v i ∈ Z d and � � Λ ∆ = x i ∈ Z i =0 i =0 11

  12. Example Λ ∆ = (0,2) v △ { 0 , (1 / 2 , 1 / 2 , 0) , (1 / 2 , 0 , 1 / 2) , (0 , 1 / 2 , 1 / 2) } = 2 ∼ � (1 / 2 , 1 / 2 , 0) , (1 / 2 , 0 , 1 / 2) � = ( Z / 2 Z ) 2 (0,0) (2,0) v v 0 1 Λ ∆ forms a finite abelian group . In this way, from a lattice simplex ∆, we can construct a finite abelian subgroup Λ ∆ of ( R / Z ) d +1 . 12

  13. FACTS (the volume of ∆) · d ! = (the order of Λ ∆ ) �� d � i =0 x i ∈ Z ≥ 0 : ( x 0 , . . . , x d ) ∈ Λ ∆ , 0 ≤ x i < 1 deg(∆) = max ∆ is NOT a lattice pyramid ⇐ ⇒ 0 ≤ ∀ i ≤ d , ∃ x ∈ Λ ∆ s.t. x i � = 0 On the other hand, from a finite abelian subgroup Λ ⊂ ( R / Z ) d +1 s.t. the sum of entries of each element in Λ is an integer. · · · · · · ( ∗ ) we can construct a lattice simplex of dim d . 13

  14. ✓ ✏ Correspondence (Batyrev–Hofscheier ’13) { lattice simplices of dim d } / ( unimod. equiv. ) 1:1 ← → { fin. abel. subgroups Λ ⊂ ( R / Z ) d +1 with ( ∗ ) } / (permute of coord.) ✒ ✑ Remark → Λ ∆ ⊂ ( R / Z ) d +1 d + 1 (dimension of ∆) ← s (degree of ∆) ← → maximum of entry sums of Λ ∆ NOT a lattice pyramid ← → 0 ≤ ∀ i ≤ d , ∃ x ∈ Λ ∆ s.t. x i � = 0 14

  15. 3. The case d + 1 = 4 s − 1 Recall For a lattice simplex ∆ of dim d with deg s , we have d + 1 ≤ f (2 s ) ≤ 4 s − 1. ✓ ✏ Our Goal Give a complete characterization of lattice simplices of d + 1 = f (2 s ) . dimension d with degree s satisfying ✿✿✿✿✿✿✿✿✿✿✿✿✿✿✿ ✒ ✑ First, we consider the case d + 1 = 4 s − 1 , which automatically implies d + 1 = f (2 s ). 15

  16. Binary Simplex Codes C ⊂ ( Z / 2 Z ) d +1 : binary simplex code ⇐ ⇒ Binary simplex code is a binary code generated by the row vectors of the matrix H ( d + 1) { column vectors of H ( d + 1) } = { T ( a 1 , . . . , a d +1 ) � = 0 : a i ∈ { 0 , 1 }} Example   1 1 1 0 1 0 0    1 1 0   H (2) = H (3) = 1 1 0 1 0 1 0    1 0 1   1 0 1 1 0 0 1 binary simplex code ⇐ ⇒ a dual code of Hamming code 16

  17. We can identify a binary code C ⊂ F d +1 as a finite abelian 2 subgroup Λ ⊂ { 0 , 1 / 2 } d +1 . We may replace 0 ∈ F 2 ← → 0 ∈ { 0 , 1 / 2 } ⊂ R / Z and 1 ∈ F 2 ← → 1 / 2 ∈ { 0 , 1 / 2 } ⊂ R / Z . Example (0,2) △ v 2 Λ ∆ = � (1 / 2 , 1 / 2 , 0) , (1 / 2 , 0 , 1 / 2) � comes from H (2) . (0,0) (2,0) v v 0 1 By Batyrev–Nill (2007), we know that this triangle is a unique lattice simplex of dim 2 with deg 1 s.t. d + 1 = 4 s − 1. 17

  18. Theorem (H. ’16) ∆ : lattice simplex of dim d with deg s satisfying d + 1 = 4 s − 1 Then s = 2 r for some r ∈ Z ≥ 0 and Λ ∆ comes from binary simplex codes . r ∈ Z ≥ 0 Proposition (H. ’16) ∆( r ) : lattice simplex of dim d with deg s = 2 r s.t. d + 1 = 4 s − 1 Then C (∆( r )) = 4 s − 1 = d + 1 . 3 3 ✓ ✏ We can see that ∆( r ) becomes a counterexample for Strong Cayley Conjecture if r ≥ 1. ✒ ✑ 18

  19. The following looks strange and unnatural... ✓ ✏ Question (Modified Strong Cayley Conjecture?) Let P be a lattice polytope of dimension d with degree s . d > 8 s − 2 ⇒ C ( P ) ≥ d + 1 − 8 s − 2 = ? 3 3 ✒ ✑ ∆( r ) satisfies this conjecture for any r ≥ 0. Remark ∆( r ) is the “most extremal” among the simplices ∆ of dim d with deg s s.t. d + 1 = f (2 s ). − → MSSC is always true for any simplices? 19

  20. 4. The case d + 1 = f (2 s ) (j.w.w. K. Kashiwabara) Recall For a lattice simplex ∆ of dim d with deg s , we have d + 1 ≤ f (2 s ) ≤ 4 s − 1. We want to give a complete characterization of lattice simplices of dim d with deg s satisfying d + 1 = f (2 s ) and compute C (∆) (for checking MSCC). ∞ � m � � By the way . . . . . . what is f ( m ) = ?? 2 ℓ ℓ =0 Example f (2) = 2 + 1 , f (3) = 3 + 1 = 4 , f (4) = 4 + 2 + 1 , f (5) = 5 + 2 + 1 , f (6) = 6 + 3 + 1 , f (7) = 7 + 3 + 1 , f (8) = 8 + 4 + 2 + 1 , f (9) = 9 + 4 + 2 + 1 . . . . . . 20

  21. Proposition f ( m ) = 2 m − ( ♯ of 1’s for the binary expansion of m ) In particular, f ( m ) = 2 m − 1 if and only if m is a power of 2. Theorem (again) (H. ’16) ∆ : lattice simplex of dim d with deg s satisfying d + 1 = 4 s − 1 s = 2 r for some r ∈ Z ≥ 0 (obvious from above Prop) and Λ ∆ Then ✿✿✿✿✿✿ comes from binary simplex codes . In particular, Λ ∆ ⊂ { 0 , 1 / 2 } d +1 . Theorem Let ∆ be a lattice simplex ∆ of dim d with deg s s.t. d + 1 = f (2 s ). ⇒ Λ ∆ ⊂ { 0 , 1 / 2 } d +1 = 21

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend