GA for straight-line grid drawings of maximal planar graphs
Author(s) -
Mohamed A. El-Sayed
Publication year - 2012
Publication title -
egyptian informatics journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.728
H-Index - 34
eISSN - 2090-4754
pISSN - 1110-8665
DOI - 10.1016/j.eij.2012.01.003
Subject(s) - grid , computer science , graph drawing , planar graph , vertex (graph theory) , planar , enhanced data rates for gsm evolution , line (geometry) , point (geometry) , algorithm , line segment , combinatorics , genetic algorithm , graph , mathematical optimization , theoretical computer science , geometry , mathematics , computer graphics (images) , artificial intelligence , machine learning
A straight-line grid drawing of a planar graph G of n vertices is a drawing of G on an integer grid such that each vertex is drawn as a grid point and each edge is drawn as a straight-line segment without edge crossings. Finding algorithms for straight-line grid drawings of maximal planar graphs (MPGs) in the minimum area is still an elusive goal. In this paper we explore the potential use of genetic algorithms to this problem and various implementation aspects related to it. Here we introduce a genetic algorithm, which nicely draws MPG of moderate size. This new algorithm draws these graphs in a rectangular grid with area ⌊2(n-1)/3⌋×⌊2(n-1)/3⌋ at least, and that this is optimal area (proved mathematically). Also, the novel issue in the proposed method is the fitness evaluation method, which is less costly than a standard fitness evaluation procedure. It is described, tested on several MPG
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom