Miro Spönemann

Drawing layered graphs with port constraints

Complex software systems are often modeled using data flow diagrams, in which nodes are connected to each other through dedicated connection points called ports. The influence a layout algorithm has on the placement of ports is determined by port constraints defined on the corresponding node. In this paper we present approaches for integrating port constraints into the layer-based approach to graph drawing pioneered by Sugiyama et al. We show how our layout algorithm, called KLay Layered, progresses from relaxed to more restricted port constraint levels as it executes, and how established algorithms for crossing minimization and edge routing can be extended to support port constraints. Compared to the previous layout algorithms supporting ports, our algorithm produces fewer edge crossings and bends and yields pleasing results. We also explain and evaluate how layout algorithms can be kept simple by using the concept of intermediate processors to structure them in a modular way. A case study integrating our layout algorithm into UC Berkeleys Ptolemy tool illustrates how KLay Layered can be integrated into Java-based applications.

Just model! — Putting automatic synthesis of node-link-diagrams into practice

Node-link-diagrams can effectively communicate information, but their creation and maintenance require a lot of manual effort. Therefore we follow the transient views approach that aims at automatically deriving high quality diagrams from arbitrary models. Besides composing diagram structures, this task involves the arrangement of the diagram elements on the canvas, and, on a finer-grained level of detail, the arrangement of the shapes (rectangles, circles, lines, etc.) that form the diagram elements. We show the feasibility of this approach by means of the Kieler Lightweight Diagrams (KLighD) framework that creates diagrams this way. We discuss our overall design objectives in terms of this framework, investigate an alternative way to shape diagram figures, and briefly demonstrate the usage of KLighD in custom modeling environments by means of a case study.

Improved layout for data flow diagrams with port constraints

The automatic generation of graphical views for data flow models and the efficient development of such models require layout algorithms that are able to handle their specific requirements. Examples include constraints on the placement of ports as well as the proper handling of nested models. We present an algorithm for laying out data flow diagrams that improves earlier approaches by reducing the number of edge crossings and bend points. We validate the quality of our algorithm with a range of models drawn from Ptolemy, a popular modeling tool for the design of embedded systems.

The open graph archive: a community-driven effort

A graphbase, a term coined by Knuth [7], is a database of graphs and computer programs that generate, analyze, manipulate, and visualize graphs. The terms graphlibrary and grapharchive are often used as synonyms for this term. Our vision is to provide an infrastructure and quality standards for a public graphbase, named the Open Graph Archive, that is accessible to researchers and other interested parties around the world via the worldwide web. This paper describes the current work undertaken towards this goal; the paper is also intended to be a call for participation since this will be a community-driven effort where most of the content will be provided by users of the system.

Port constraints in hierarchical layout of data flow diagrams

We present a new application for graph drawing in the context of graphical model-based system design, where manual placing of graphical items is still state-of-the-practice. The KIELER framework aims at improving this by offering novel user interaction techniques, enabled by automatic layout of the diagrams. In this paper we present extensions of the well-known hierarchical layout approach, originally suggested by Sugiyama et al. [22], to support port constraints, hyperedges, and compound graphs in order to layout diagrams of data flow languages. A case study and experimental results show that our algorithm is well suited for application in interactive user interfaces.

Evolutionary Meta Layout of Graphs

A graph drawing library is like a toolbox, allowing experts to select and configure a specialized algorithm in order to meet the requirements of their diagram visualization application. However, without expert knowledge of the algorithms the potential of such a toolbox cannot be fully exploited. This gives rise to the question whether the process of selecting and configuring layout algorithms can be automated such that good layouts are produced. In this paper we call this kind of automation “meta layout.” We propose a genetic representation that can be used in meta heuristics for meta layout and contribute new metrics for the evaluation of graph drawings. Furthermore, we examine the use of an evolutionary algorithm to search for optimal solutions and evaluate this approach both with automatic experiments and a user study.

A Generalization of the Directed Graph Layering Problem

The Directed Layering Problem (DLP) solves a step of the widely used layer-based approach to automatically draw directed acyclic graphs. To cater for cyclic graphs, usually a preprocessing step is used that solves the Feedback Arc Set Problem (FASP) to make the graph acyclic before a layering is determined.

Crossing minimization and layouts of directed hypergraphs with port constraints

Many practical applications for drawing graphs are modeled by directed graphs with domain specific constraints. In this paper, we consider the problem of drawing directed hypergraphs with (and without) port constraints, which cover multiple real-world graph drawing applications like data flow diagrams and electric schematics. Most existing algorithms for drawing hypergraphs with port constraints are adaptions of the framework originally proposed by Sugiyama et al. in 1981 for simple directed graphs. Recently, a practical approach for upward crossing minimization of directed graphs based on the planarization method was proposed [7]. With respect to the number of arc crossings, it clearly outperforms prior (mostly layering-based) approaches. We show how to adopt this idea for hypergraphs with given port constraints, obtaining an upward-planar representation (UPR) of the input hypergraph where crossings are modeled by dummy nodes. Furthermore, we present the new problem of computing an orthogonal upward drawing with minimal number of crossings from such an UPR, and show that it can be solved efficiently by providing a simple method.

Generalized Layerings for Arbitrary and Fixed Drawing Areas

The Directed Layering Problem (DLP) solves a step of the widely used layer-based approach to automatically draw directed acyclic graphs. To cater for cyclic graphs, usually a preprocessing step is used that solves the Feedback Arc Set Problem (FASP) to make the graph acyclic before a layering is determined. Here we present the Generalized Layering Problem (GLP), which solves the combination of DLP and FASP simultaneously, allowing general graphs as input. We present an integer programming model and a heuristic to solve the NP-complete GLP and perform thorough evaluations on different sets of graphs and with different implementations for the steps of the layer-based approach. We observe that GLP reduces the number of dummy nodes significantly, can produce more compact drawings, and improves on graphs where DLP yields poor aspect ratios. The drawings resulting from GLP also turn out to be more suitable for making the best possible use of a given drawing area. However, we show that a specialized variant of GLP can yield considerable improvements w. r. t. this particular optimization goal.

Counting Crossings for Layered Hypergraphs

Orthogonally drawn hypergraphs have important applications, e. g. in actor-oriented data flow diagrams for modeling complex software systems. Graph drawing algorithms based on the approach by Sugiyama et al. place nodes into consecutive layers and try to minimize the number of edge crossings by finding suitable orderings of the nodes in each layer. With orthogonal hyperedges, however, the exact number of crossings is not determined until the edges are actually routed in a later phase of the algorithm, which makes it hard to evaluate the quality of a given node ordering beforehand. In this paper, we present and evaluate two cross counting algorithms that predict the number of crossings between orthogonally routed hyperedges much more accurately than the traditional straight-line method.