Periodic Parallelogram Polyominoes P. Laborde-Zubieta joint work with A. Boussicault LaBRI - Université de Bordeaux GASCOM 2016 - 04/06/2016 1 / 17 Outline Definitions Decomposition of PPPs in ordered trees 2 / 17 Parallelogram Polyominoes A parallelogram polyomino is a set of edge connected cells, whose boundary can be decomposed in two paths : a lower path and an upper path. They are made of north and east steps, and they meet only at their starting and ending point. 3 / 17 Periodic Parallelogram Polyominoes A periodic parallelogram polyomino (PPP) is a parallelogram polyomino with a marked cell in the rightmost column. 4 / 17 Periodic Parallelogram Polyominoes A periodic parallelogram polyomino (PPP) is a parallelogram polyomino with a marked cell in the rightmost column. The marking corresponds to the location where we “glue” the leftmost and the rightmost columns. The number of cells above the marked cell, included, is lower or equal to the number of cells in the leftmost column. 4 / 17 Periodic Parallelogram Polyominoes A periodic parallelogram polyomino (PPP) is a parallelogram polyomino with a marked cell in the rightmost column. The marking corresponds to the location where we “glue” the leftmost and the rightmost columns. The number of cells above the marked cell, included, is lower or equal to the number of cells in the leftmost column. 4 / 17 Periodic Parallelogram Polyominoes A periodic parallelogram polyomino (PPP) is a parallelogram polyomino with a marked cell in the rightmost column. The marking corresponds to the location where we “glue” the leftmost and the rightmost columns. The number of cells above the marked cell, included, is lower or equal to the number of cells in the leftmost column. 4 / 17 Height, width, semi-perimeter I Height : the number of rows below the marked row I Width : the number of columns I Semi-perimeter : the height + the width 1 2 4 5 4 3 5 6 1 8 7 2 3 4 5 5 / 17 Rotation and strips Making a rotation to a PPP means, putting the leftmost column in the rightmost position and updating the marking. rotation 6 / 17 Rotation and strips Making a rotation to a PPP means, putting the leftmost column in the rightmost position and updating the marking. rotation The rotation induces a partitioning of PPPs in equivalent classes, we call them strips. Since the rotation does not modify the semi-perimeter, we call semi-perimeter of a strip, the semi-perimeter of one of its PPP. 6 / 17 Plan Definitions Decomposition of PPPs in ordered trees 7 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 4 2 5 4 3 6 1 8 7 2 3 4 5 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 4 2 5 4 3 6 1 8 7 2 3 4 5 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 2 4 5 4 3 5 6 1 8 7 2 3 4 5 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Underlying graph of a PPP The blue and red labels are there to help the understanding, they are not necessary. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 The sons of a vertex are ordered according to their order in the PPP. The number of vertices of the graph is equal to the semi-perimeter of the PPP. The graph is bipartite, hence each cycle is of even size. All the PPP of a same Strip have the same underlying graph. 8 / 17 Description of graphs Since the number of vertices is equal to the number of edges in each connected component, by the following formula, cyclomatic number − 1 = #edges − #vertices, the cyclomatic number of each connected component is equal to 1. 9 / 17 Description of graphs Since the number of vertices is equal to the number of edges in each connected component, by the following formula, cyclomatic number − 1 = #edges − #vertices, the cyclomatic number of each connected component is equal to 1. Moreover, there is at least one cycle per connected component. 9 / 17 Description of graphs Since the number of vertices is equal to the number of edges in each connected component, by the following formula, cyclomatic number − 1 = #edges − #vertices, the cyclomatic number of each connected component is equal to 1. Moreover, there is at least one cycle per connected component. Hence, each connected component is a cycle with two ordered trees attached to each vertex of the cycle, one “inside” and one “outside”. 9 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 3 5 6 5 1 8 7 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 3 5 6 5 1 8 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 3 5 6 5 1 8 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 5 6 5 1 8 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 5 6 5 1 8 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 5 6 5 1 2 3 4 5 2 4 3 8 4 2 3 6 5 7 How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. How do we reconstruct the leaf ? We have two possibilities. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. To be able to reconstruct back, we need to mark the vertex corresponding to the leftmost column. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 2 1 4 5 4 5 6 5 1 2 3 4 5 2 4 3 8 2 3 4 6 5 7 To be able to reconstruct back, we need to mark the vertex corresponding to the leftmost column. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 1 2 4 5 4 5 5 6 1 4 4 3 2 5 2 8 2 3 4 6 5 7 To be able to reconstruct back, we need to mark the vertex corresponding to the leftmost column. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 1 2 4 5 4 5 5 6 1 4 4 3 2 5 2 8 2 3 4 6 5 7 To be able to reconstruct back, we need to mark the vertex corresponding to the leftmost column. 10 / 17 Pruning Pruning a PPP consists in removing recursively the rows and the columns corresponding to leaves in the underlying grah. 1 1 1 2 4 4 5 5 6 1 4 4 3 2 5 2 8 2 3 4 6 5 7 To be able to reconstruct back, we need to mark the vertex corresponding to the leftmost column. 10 / 17 Trunk PPP + intrinsic thickness After pruning a PPP, we will always get a PPP of the following shape. Such a PPP is called a trunk PPP. We define the intrinsic thickness of a trunk PPP to be the height of one of its column minus 1. A trunk PPP is completely determined by its number of columns and its intrinsic thickness. The intrinsic thickness of a general PPP, is the intrinsic thickness of the trunk PPP obtained after pruning it. 11 / 17 Trunk PPP + intrinsic thickness After pruning a PPP, we will always get a PPP of the following shape. Such a PPP is called a trunk PPP. We define the intrinsic thickness of a trunk PPP to be the height of one of its column minus 1. A trunk PPP is completely determined by its number of columns and its intrinsic thickness. The intrinsic thickness of a general PPP, is the intrinsic thickness of the trunk PPP obtained after pruning it. 11 / 17 Trunk PPP + intrinsic thickness After pruning a PPP, we will always get a PPP of the following shape. Such a PPP is called a trunk PPP. We define the intrinsic thickness of a trunk PPP to be the height of one of its column minus 1. A trunk PPP is completely determined by its number of columns and its intrinsic thickness. The intrinsic thickness of a general PPP, is the intrinsic thickness of the trunk PPP obtained after pruning it. 11 / 17 Trunk PPP + intrinsic thickness After pruning a PPP, we will always get a PPP of the following shape. Such a PPP is called a trunk PPP. We define the intrinsic thickness of a trunk PPP to be the height of one of its column minus 1. A trunk PPP is completely determined by its number of columns and its intrinsic thickness. The intrinsic thickness of a general PPP, is the intrinsic thickness of the trunk PPP obtained after pruning it. 11 / 17 Underlying graph of a trunk PPP The underlying graph of a trunk PPP is a disjoint union of cycles of the same even size. It is not injective, two trunk PPP can have the same underlying graph. 1 2 4 3 3 3 2 1 4 4 1 2 3 4 1 2 1 2 3 4 3 4 4 3 2 1 4 3 2 1 12 / 17 Decomposition of a PPP in trees 1 1 2 3 4 1 2 3 4 2 1 2 3 4 3 4 13 / 17 Decomposition of a PPP in trees 1 1 2 3 4 1 2 3 4 1 2 3 4 3 4 ! tint = 2 2 13 / 17 Decomposition of a PPP in trees 1 1 2 3 4 1 2 3 4 1 2 3 4 3 4 ! tint = 2 2 13 / 17 Decomposition of a PPP in trees 1 2 4 1 2 3 4 1 2 3 ! tint = 2 1 2 3 4 3 4 13 / 17 PPPs : Bijection, enumeration Theorem PPPs (except rectangular shaped ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a list of 4-tuples of bicolored ordered trees such that: I each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees, I a black vertex is marked in the first 4-tuple. 14 / 17 PPPs : Bijection, enumeration Theorem PPPs (except rectangular shaped ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a list of 4-tuples of bicolored ordered trees such that: I each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees, I a black vertex is marked in the first 4-tuple. The sequence enumerating, with respect to the semi-perimeter, the PPPs of fixed intrinsic thickness, are counted by the sequence 2n+1 n (4 − n )n>1 (A008549). 14 / 17 PPPs : Bijection, enumeration Theorem PPPs (except rectangular shaped ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a list of 4-tuples of bicolored ordered trees such that: I each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees, I a black vertex is marked in the first 4-tuple. The sequence enumerating, with respect to the semi-perimeter, the PPPs of fixed intrinsic thickness, are counted by the sequence 2n+1 n (4 − n )n>1 (A008549). It counts : I “The sum of the areas under all Dyck excursions of length 2n.” I “Number of inversions in all 321-avoiding permutations of [n + 1].” 14 / 17 PPPs : Bijection, generating function. Theorem PPPs (except rectangular shaped ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a list of 4-tuples of bicolored ordered trees such that: I each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees, I a black vertex is marked in the first 4-tuple. 15 / 17 PPPs : Bijection, generating function. Theorem PPPs (except rectangular shaped ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a list of 4-tuples of bicolored ordered trees such that: I each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees, I a black vertex is marked in the first 4-tuple. A• (z• , z◦ ) = 1 1 et A◦ (z• , z◦ ) = . 1 − z◦ A◦ 1 − z• A• PPP(z• , z◦ ) = z• ∂z• z• z◦ A• (z• , z◦ )2 A◦ (z• , z◦ )2 . 1 − z• z◦ A• (z• , z◦ )2 A◦ (z• , z◦ )2 15 / 17 PPPs : Bijection, generating function. Theorem PPPs (except rectangular shaped ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a list of 4-tuples of bicolored ordered trees such that: I each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees, I a black vertex is marked in the first 4-tuple. A• (z• , z◦ ) = 1 1 et A◦ (z• , z◦ ) = . 1 − z◦ A◦ 1 − z• A• PPP(z• , z◦ ) = PPP(z, z) = z• ∂z• z• z◦ A• (z• , z◦ )2 A◦ (z• , z◦ )2 . 1 − z• z◦ A• (z• , z◦ )2 A◦ (z• , z◦ )2 zC (z)2 z , where C(z) = . 1 − 4z 1 − C(z) 15 / 17 Strips : Bijection, generating function Theorem Strips (except rectangular ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a cycle of 4-tuples of bicolored ordered trees such that each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees. 16 / 17 Strips : Bijection, generating function Theorem Strips (except rectangular ones) are in bijection with pairs consisting of a positive integer corresponding to the intrinsic thickness and a cycle of 4-tuples of bicolored ordered trees such that each 4-tuple is composed of 2 black rooted trees and 2 white rooted trees. By Pólya’s theory, the generating function of Strips, with respect to the semi-perimeter, with a fixed intrinsic thickness: S(z) = − X ϕ(i) i>1 i log(1 − z 2i A(z i )4 ), where ϕ is the Euler phi function and A(z) = 1 . 1 − zA(z) 16 / 17 Thank you. 17 / 17
© Copyright 2026 Paperzz