combinatorics of the double dimer model
play

Combinatorics of the Double-Dimer Model Helen Jenne University of - PowerPoint PPT Presentation

Combinatorics of the Double-Dimer Model Helen Jenne University of Oregon Dimers in Combinatorics and Cluster Algebras 2020 August 10, 2020 This talk is being recorded 1 / 25 Outline Kuo Condensation 1 Main Result: Double-Dimer Condensation


  1. Combinatorics of the Double-Dimer Model Helen Jenne University of Oregon Dimers in Combinatorics and Cluster Algebras 2020 August 10, 2020 This talk is being recorded 1 / 25

  2. Outline Kuo Condensation 1 Main Result: Double-Dimer Condensation 2 Ideas of Proof 3 Non-tripartite pairings 4 2 / 25

  3. Kuo condensation Today G = ( V 1 , V 2 , E ) is a finite bipartite planar graph. Let Z D ( G ) denote the partition function . Z D ( G ) = xyz + x + z y x z Theorem (Kuo04, Theorem 5.1) Let vertices a , b , c , and d appear in a cyclic order on a face of G . If a , c 2 V 1 and b , d 2 V 2 , then Z D ( G ) Z D ( G � { a , b , c , d } )= Z D ( G � { a , b } ) Z D ( G � { c , d } )+ Z D ( G � { a , d } ) Z D ( G � { b , c } ) a d c b a a a d d d c c c b b b 3 / 25

  4. Kuo Condensation Theorem (Kuo04, Theorem 5.1) Let vertices a , b , c , and d appear in a cyclic order on a face of G . If a , c 2 V 1 and b , d 2 V 2 , then Z D ( G ) Z D ( G � { a , b , c , d } )= Z D ( G � { a , b } ) Z D ( G � { c , d } )+ Z D ( G � { a , d } ) Z D ( G � { b , c } ) Examples of non-bijective proofs: Fulmek, Graphical condensation, overlapping Pfa ffi ans and superpositions of Matchings Speyer, Variations on a theme of Kasteleyn, with Application to the TNN Grassmannian Theorem (Desnanot-Jacobi identity/Dodgson condensation) det( M ) det( M i , j i ) det( M j j ) � det( M j i , j ) = det( M i i ) det( M i j ) M j i is the matrix M with the i th row and the j th column removed. 4 / 25

  5. Applications of Kuo’s work Tiling enumeration New proof of MacMahon’s product formula for the generating function for plane partitions that are subsets of an r ⇥ s ⇥ t box. Cluster algebras (LM17) Toric cluster variables for the quiver associated to the cone of the del Pezzo surface of degree 6 Main result. An analogue of Kuo’s theorem for double-dimer configs. Application: A problem in Donaldson-Thomas theory and Pandharipande-Thomas theory (joint work with Ben Young and Gautam Webb) 5 / 25

  6. Double-dimer configurations N is a set of special vertices called nodes on the outer face of G . Definition (Double-dimer configuration on ( G , N )) 7 6 5 Configuration of ` disjoint loops 8 Doubled edges 4 Paths connecting nodes in pairs Its weight is the product of its edge weights ⇥ 2 ` 1 2 3 7 6 5 7 6 5 8 8 = + 4 4 1 1 2 3 2 3 6 / 25

  7. Tripartite pairings Definition (Tripartite pairing) A planar pairing � of N is tripartite if the nodes can be divided into  3 sets of circularly consecutive nodes so that no node is paired with a node in the same set. 1 1 2 12 2 12 3 3 11 4 4 10 5 11 5 9 6 7 8 9 10 6 8 7 Tripartite Not tripartite We often color the nodes in the sets red, green, and blue, in which case � has no monochromatic pairs. Dividing nodes into three sets R , G , and B defines a tripartite pairing. 7 / 25

  8. Main Result Z DD ( G , N ) denotes the weighted sum of all DD config with pairing � . � Theorem (J.) Divide N into sets R , G , and B and let � be the corr. tripartite pairing. Let x , y , w , v 2 N such that x < w 2 V 1 and y < v 2 V 2 . If { x , y , w , v } contains at least one node of each RGB color and x , y , w , v appear in cyclic order then Z DD ( G , N ) Z DD σ xywv ( G , N � { x , y , w , v } ) = σ Z DD σ xy ( G , N � { x , y } ) Z DD σ wv ( G , N � { w , v } ) + Z DD σ xv ( G , N � { x , v } ) Z DD σ wy ( G , N � { w , y } ) Example. Z DD σ ( N ) Z DD σ 1258 ( N − 1 , 2 , 5 , 8) = Z DD σ 12 ( N − 1 , 2) Z DD σ 58 ( N − 5 , 8)+ Z DD σ 18 ( N − 1 , 8) Z DD σ 25 ( N − 2 , 5) 7 6 5 7 6 7 6 5 7 6 7 6 5 7 6 8 8 8 4 4 4 4 4 4 1 1 1 2 3 3 3 2 3 2 3 3 8 / 25

  9. Main Result Z DD ( G , N ) denotes the weighted sum of all DD config with pairing � . � Theorem (J.) Divide N into sets R , G , and B and let � be the corr. tripartite pairing. Let x , y , w , v 2 N such that x < w 2 V 1 and y < v 2 V 2 . If { x , y , w , v } contains at least one node of each RGB color and x , y , w , v appear in cyclic order then Z DD ( G , N ) Z DD σ xywv ( G , N � { x , y , w , v } ) = σ Z DD σ xy ( G , N � { x , y } ) Z DD σ wv ( G , N � { w , v } ) + Z DD σ xv ( G , N � { x , v } ) Z DD σ wy ( G , N � { w , y } ) Example. Z DD σ ( N ) Z DD σ 1258 ( N − 1 , 2 , 5 , 8) = Z DD σ 12 ( N − 1 , 2) Z DD σ 58 ( N − 5 , 8)+ Z DD σ 18 ( N − 1 , 8) Z DD σ 25 ( N − 2 , 5) 7 6 5 7 6 7 6 5 7 6 7 6 5 7 6 8 8 8 4 4 4 4 4 4 1 1 1 2 3 3 3 2 3 2 3 3 We only need the two nodes of the same RGB color to be opposite in BW color. 8 / 25

  10. Corollaries Theorem (Kuo04, Theorem 5.1) a d Let vertices a , b , c , and d appear in a cyclic order on a face of c G . If a , c 2 V 1 and b , d 2 V 2 , then b Z D ( G ) Z D ( G � { a , b , c , d } )= Z D ( G � { a , b } ) Z D ( G � { c , d } )+ Z D ( G � { a , d } ) Z D ( G � { b , c } ) Theorem (J.) Let x , y , w , v 2 N such that x < w 2 V 1 and 7 6 5 y < v 2 V 2 . If { x , y , w , v } contains at least one node of each RGB color and the two nodes of the same RGB 8 color are opposite in BW color then 4 Z DD ( G , N ) Z DD σ xywv ( G � { x , y , w , v } , N � { x , y , w , v } ) = σ Z DD σ xy ( G � { x , y } , N � { x , y } ) Z DD 1 2 3 σ wv ( G � { w , v } , N � { w , v } ) + Z DD σ xv ( G � { x , v } , N � { x , v } ) Z DD σ wy ( G � { w , y } , N � { w , y } ) 9 / 25

  11. Corollaries Theorem (Kuo04, Theorem 5.1) a d Let vertices a , b , c , and d appear in a cyclic order on a face of c G . If a , c 2 V 1 and b , d 2 V 2 , then b Z D ( G ) Z D ( G � { a , b , c , d } )= Z D ( G � { a , b } ) Z D ( G � { c , d } )+ Z D ( G � { a , d } ) Z D ( G � { b , c } ) Theorem (J.) Let x , y , w , v 2 N such that x < w 2 V 1 and 7 6 5 y < v 2 V 2 . If { x , y , w , v } contains at least one node of each RGB color and the two nodes of the same RGB color are opposite in BW color then 4 Z DD ( G , N ) Z DD σ xywv ( G � { x , y , w , v } , N � { x , y , w , v } ) = σ Z DD σ xy ( G � { x , y } , N � { x , y } ) Z DD 2 3 σ wv ( G � { w , v } , N � { w , v } ) + Z DD σ xv ( G � { x , v } , N � { x , v } ) Z DD σ wy ( G � { w , y } , N � { w , y } ) 9 / 25

  12. Corollaries Theorem (Kuo04, Theorem 5.2) a d Let vertices a , c , b , and d appear in a cyclic order on a face of G . If a , c 2 V 1 and b , d 2 V 2 , then c b Z D ( G ) Z D ( G � { a , b , c , d } )= Z D ( G � { a , d } ) Z D ( G � { b , c } ) � Z D ( G � { a , b } ) Z D ( G � { c , d } ) Theorem (J.) Let x , y , w , v 2 N such that x < w 2 V 1 and y < v 2 V 2 . If { x , y , w , v } contains at least one node of each RGB color and the two nodes of the same RGB color are the same in BW color then Z DD ( G , N ) Z DD σ xywv ( G � { x , y , w , v } , N � { x , y , w , v } ) = σ Z DD σ xy ( G � { x , y } , N � { x , y } ) Z DD σ wv ( G � { w , v } , N � { w , v } ) � Z DD σ xv ( G � { x , v } , N � { x , v } ) Z DD σ wy ( G � { w , y } , N � { w , y } ) 10 / 25

  13. Corollaries Theorem (Kuo04, Theorem 5.2) a d Let vertices a , c , b , and d appear in a cyclic order on a face of G . If a , c 2 V 1 and b , d 2 V 2 , then c b Z D ( G ) Z D ( G � { a , b , c , d } )= Z D ( G � { a , d } ) Z D ( G � { b , c } ) � Z D ( G � { a , b } ) Z D ( G � { c , d } ) Theorem (J.) Let x , y , w , v 2 N such that x < w 2 V 1 and y < v 2 V 2 . If { x , y , w , v } contains at least one node of each RGB color and the two nodes of the same RGB color are the same in BW color then Z DD ( G , N ) Z DD σ xywv ( G � { x , y , w , v } , N � { x , y , w , v } ) = σ Z DD σ xy ( G � { x , y } , N � { x , y } ) Z DD σ wv ( G � { w , v } , N � { w , v } ) � Z DD σ xv ( G � { x , v } , N � { x , v } ) Z DD σ wy ( G � { w , y } , N � { w , y } ) 10 / 25

  14. Background: Double-dimer pairing probabilities 1 2 6 ⇣ 1 3 5 ⌘ b Pr = X 1 , 4 X 2 , 5 X 3 , 6 + X 1 , 2 X 3 , 4 X 5 , 6 2 4 6 3 5 4 ⇣ 1 3 5 7 ⌘ b Pr = X 1 , 8 X 3 , 4 X 5 , 2 X 7 , 6 � X 1 , 4 X 3 , 8 X 5 , 2 X 7 , 6 + X 1 , 6 X 3 , 4 X 5 , 8 X 7 , 2 8 4 2 6 � X 1 , 8 X 3 , 6 X 5 , 2 X 7 , 4 � X 1 , 4 X 3 , 6 X 5 , 8 X 7 , 2 + X 1 , 6 X 3 , 8 X 5 , 2 X 7 , 4 Definition (KW11a) Z D ( G BW ) Z D ( G BW ) , where G BW ✓ G only contains nodes that are black and X i , j = i , j odd or white and even. 4 1 1 4 1 4 1 4 3 2 2 2 3 2 G BW G BW G = G BW G BW G 1 , 2 2 , 4 11 / 25

  15. X i , j = 0 if i and j have the same parity 1 1 1 2 8 2 8 2 8 3 7 3 7 3 7 4 6 4 6 4 6 ⇣ 1 3 5 7 ⌘ 5 5 5 b Pr = X 1 , 8 X 3 , 4 X 5 , 2 X 7 , 6 � X 1 , 4 X 3 , 8 X 5 , 2 X 7 , 6 + X 1 , 6 X 3 , 4 X 5 , 8 X 7 , 2 8 4 2 6 � X 1 , 8 X 3 , 6 X 5 , 2 X 7 , 4 � X 1 , 4 X 3 , 6 X 5 , 8 X 7 , 2 + X 1 , 6 X 3 , 8 X 5 , 2 X 7 , 4 Each term in b Pr( � ) is of the form Q X ⌧ := X i , j , where ⌧ is an odd-even pairing. ( i , j ) 2 ⌧ Kenyon and Wilson made a simplifying assumption that all nodes are black and odd or white and even. Theorem (KW11a, Theorem 1.3) b Pr ( � ) is an integer-coe ff homogeneous polynomial in the quantities X i , j 12 / 25

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