Scalable Drawing of Nested Directed Acyclic Graphs With Gates and Ports

Abstract

Graphs are a popular concept throughout all of computer science, from networking through data management to machine learning. This has given rise to the problem of automatic graph drawing, yet efficient and aesthetically-pleasing graph layouting is as diverse as the underlying graph characteristics. In this thesis we deal with graphs that not only can be nested in two different ways, but in which nodes also have ports. We derive two alternative methods from state-of-the-art approaches. Our focus is on both the quality of the drawings and the performance of the layouters. We then evaluate and compare our algorithms on a total of 40 real-world graphs. We find that at least one of them, a level-based algorithm, is able to generate adequate drawings in a scalable manner.