Visualization is often invaluable to understand the behavior of optimization algorithms, identify their bottlenecks or pathological behaviors, and suggest remedies. Yet developing visualizations is a tedious activity requiring significant time and expertise. This paper presents a framework for the visualization of constraint-based local search (CBLS). Given a high-level model and a declarative visualization specification, the CBLS visualizer systematically produces animations to visualize constraints and objectives, violations, and conflicts, as well as the temporal behavior of these measures. The visualization specification is composed of a triple (what, where, how) indicating what to display, where, and how. The visualizer architecture is compositional and extensible, and focuses almost exclusively on static aspects, the dynamic aspects being automated by invariants. The paper highlights various functionalities of the visualizer and describes a blueprint for its implementation.
展开▼
