Presentation is loading. Please wait.

Presentation is loading. Please wait.

LaBRI, Université Bordeaux I

Similar presentations


Presentation on theme: "LaBRI, Université Bordeaux I"— Presentation transcript:

1 LaBRI, Université Bordeaux I
STACS’03 An Information-Theoretic Upper Bound of Planar Graphs Using Triangulation I will be talking to you about An Information-Theoretic Upper Bound of Planar Graphs Using Triangulation. This work has been completed by myself, Bertrand Le Saëc and Mohamed Mosbah. Nicolas Bonichon, Cyril Gavoille & Nicolas Hanusse

2 Planar Graphs (unlabeled)
Maximal planar graph (or triangulation) Plane embedding (or plane graph) First let me define the graphs we will consider in this talk. A planar graph is a graph with can be embedded in the plane. Here is an exemple of a planar graph. A planar graph G is called maximal if no edges can be added to G without breaking the planarity property. Here is an exemple of maximal plane graph. A plane embedding or plane graph for short, is a graph where the incident edges of each node are ordered. This two plane graphs are two different plane embeddings of the same planar graph.

3 p(n) = number of n-node planar graphs
Our Problem How many bits are needed to encode a planar graph with n nodes? What is the number of edges of a uniformly random planar graph? p(n) = number of n-node planar graphs p(n)? The questions that we would like to answer are the following ones: How many planar graphs with n nodesexists ? How many bits are needed to encode a planar graph. Note that this two question are strongly connected. Indeed, if any object of Pn can be encoded with alpha-n bits, then the the size of Pn is no bigger thant 2 power alpha n. If we do not constrain the complexity of the encoding algorithm, the reserve is also true. The last question we would like to answer is the following: how many edges a typical plane graph has ?

4 Related Works Encoding (n = number of nodes; m = number of edges)
[Turan, 84]: 4m bits [Keeler, Westbrook, 95]: 3.58m bits [Munro, Raman, 97]: 2m + 8n bits [Chiang, Lin, Lu, 01]: 4m/3 + 5n bits Number of planar graphs [Osthus, Prömel, Taraz, 02]:  25.22n [Bousquet-Mélou, 02]:  n [Bender, Gao, Wormald, 99]:  24.71n For triangulation [Tutte, 62]: 23.24n Number of edges (for almost all planar graphs) [Gerke, McDiamid, 01]: m  2.69n For labeled [Osthus, Prömel, Taraz, 02]: 1.85n  m  2.56n Several results gives partial answer to theses questions. Succinct representation of planar graphshas a long history. Considering the encoding problem, several algorithms have been proposed with different efficiency. For the bounds on the enumeration of planar graphs several results have also been given.

5 Encoding Scheme of a Planar Graph
Embed the graph G Triangulate the graph G Encode the triangulation Encode the edges to remove. Coding size: 3.24n + 3n = 6.24n Ideas: Compute a good embedding of G: well-orderly embedding Compute a good triangulation of G: super-triangulation

6 Realizer (or Schnyder trees) [Schnyder, 89]
R=(T0,T1,T2) is a realizer of a maximal plane graph G if: T0,T1,T2 make a partition of internal edges of G. For each internal node v: Example : v Thm [Schnyder, 89]: R=(T0,T1,T2) can be computed in linear time

7 Structure of the Realizers of G
cw-triangle ccw-triangle Thm [Ossona de Mendez, 94]: The realizers of G is a distributive lattice. The minimal realizer is the unique realizer of G without any cw-triangle.

8 Super-Triangulation S=(T0,T1,T2) is a super-triangulation of a planar graph G if: V(S) = V(G) and E(G)  E(S) S is a minimal realizer T0  E(G) If v is an inner node of T2 then (v, P1(v))  E(G) 1 3 6 2 8 7 5 4 1 3 6 2 8 7 5 4 Thm: Every connected planar graph G has a super-triangulation S that can be computed in O(n) time.

9 Encoding a Planar Graph with a Super-Triangulation
Planar graph = super-triangulation - {missing edges (green + some red)} Coding: 1

10 Coding a Planar Graph with a Super-Triangulation
Planar graph = super-triangulation - {missing edges (green + some red)} Decoding: 1

11 Length Coding Analysis
Data structure: 7 binary strings with different density of “1”. 1 binary string for the missing edges. Each string is compressed considering its density Thm: Planar graph encoding 5.03n bits (3.37n bits for the super-triangulation)

12 Enumeration Thm: Let p(n) be the number of planar graph with n nodes.
p(n)  n Thm: For almost all planar graphs, the number of edges m is: 1.70n  m  2.54n (Previously: ??  m  2.69n )

13 Conclusion Results: Conjecture: p(n)  binomial(5n,2n)  24.85n
An explicit linear time and space algorithm to encode a planar graph with 5.03n bits. A new upper-bound on the number of planar graphs: p(n)  n new bounds on the typical number of edges: 1.70n  m  2.54n Conjecture: p(n)  binomial(5n,2n)  24.85n find a better encoding of the super-triangulation find a better embedding


Download ppt "LaBRI, Université Bordeaux I"

Similar presentations


Ads by Google