Goals and results

The goal of this project was to find a way to visualize the structure of large hierarchical models. This structure is usually dynamically generated during an initialization phase, but depending on the model it might even be modified during execution. The structure, its attributes and changes over time are coded in the model. In order to check it, a visual inspection may be useful. Thus we integrated the point-based visualization into JAMES II for browsing snapshots of a model’s structure and filtering it according to given node attributes. In contrast to the CGV-integration, this tool is specialized towards the representation of hierarchies and thus eases browsing and checking hierarchical models, such as ML-DEVS models. Due to their highly regular structure, deviations from this regularity (possible errors in the model) show up as asymmetries in the point-based view and can be perceived and investigated easily even for larger models.

Figure: The JAMES II framework showing a point-based layout of a hierarchical model with 300,000 nodes.

Interdisciplinarity

For this project knowledge about the structure of models given in a certain modeling formalism and knowledge about the graphical representations of graphs was necessary. Thus we tightly interacted in order to tie together the internal data structures of both and send update events back and forth between both systems.