Tulip 5

data visualization. The second package is a complete OpenGL rendering engine (Tulip OGL) tailored for information visualization techniques. The third package is a library of GUI components created using the Qt library (Tulip GUI). The fourth package is a complete binding of the C++ API to Python. It makes the usage of Tulip inside Python code straightforward. Finally, Tulip software is an application where users can embed their algorithm, visualization technique, or complete information visualization pipeline. Figure 2 summarized the connections between these different libraries. Fig. 2. Tulip architecture overview. The Tulip framework consists of four packages. Tulip core provides an efficient data structures for relational data. Tulip graphics is a complete OpenGL rendering engine. Tulip GUI is a collection of widgets built on top of Qt library for the purpose of information visualization. Finally, Tulip software is an application for embedding algorithms, visualization techniques/interaction, and complete information visualization systems. All these packages can be dynamically extended through the plugin architecture of Tulip. 3 Historical Background Developing a framework over an extended period of time often means being compared to or challenged by competitor systems and libraries. This section presents a subset of the libraries that are closest in spirit to our work. We briefly discuss the philosophy or underlying principles of each, contrasting them to Tulip.