No abstract

Readability criteria, such as distance or neighborhood preservation, are often used to optimize node-link representations of graphs to enable the comprehension of the underlying data. With few exceptions, graph drawing algorithms typically optimize one such criterion, usually at the expense of others. We propose a layout approach, Graph Drawing via Gradient Descent, (GD) 2 , that can handle multiple readability criteria. (GD) 2 can optimize any criterion that can be described by a smooth function. If the criterion cannot be captured by a smooth function, a non-smooth function for the criterion is combined with another smooth function, or auto-differentiation tools are used for the optimization. Our approach is flexible and can be used to optimize several criteria that have already been considered earlier (e.g., obtaining ideal edge lengths, stress, neighborhood preservation) as well as other criteria which have not yet been explicitly optimized in such fashion (e.g., vertex resolution, angular resolution, aspect ratio). We provide quantitative and qualitative evidence of the effectiveness of (GD) 2 with experimental data and a functional prototype: http://hdc.cs.arizona.edu/ ~mwli/graph-drawing/.

Figure 1: CFGExplorer helps researchers analyze programs and correlate control flow (a) and trace data (b). We designed a domain-specific layout approach to elucidate loop structure (orange). Users can animate execution using the linked blue gradient (c). AbstractTo develop new compilation and optimization techniques, computer scientists frequently consult program analysis artifacts such as control flow graphs (CFGs) and traces of executed instructions. A CFG is a directed graph representing possible execution paths in a program. CFGs are commonly visualized as node-link diagrams while traces are commonly viewed in raw text format. Visualizing and exploring CFGs and traces is challenging because of the complexity and specificity of the operations researchers perform. We present a design study where we collaborate with computer scientists researching dynamic binary analysis and compilation techniques. The research group primarily employs CFGs and traces to reason about and develop new algorithms for program optimization and parallelization. Through questionnaires, interviews, and a year-long observation, we analyzed their use of visualization, noting that the tasks they perform match common subroutines they employ in their techniques. Based on this task analysis, we designed CFGExplorer, a visual analytics system that supports computer scientists with interactions that are integrated with the program structure. We developed a domain-specific graph modification to generate graph layouts that reflect program structure. CFGExplorer incorporates structures such as functions and loops, and uses the correspondence between CFGs and traces to support navigation. We further augment the system to highlight the output of program analysis techniques, facilitating exploration at a higher level. We evaluate the tool through guided sessions and semi-structured interviews as well as deployment. Our collaborators have integrated CFGExplorer into their workflow and use it to reason about programs, develop and debug new algorithms, and share their findings.

No abstract

No abstract