Integrated Performance Monitoring of a Cosmology Application on Leading HEC Platforms
ICPP '05 Proceedings of the 2005 International Conference on Parallel Processing
The Grammar of Graphics (Statistics and Computing)
The Grammar of Graphics (Statistics and Computing)
Security Data Visualization
Designing semantic substrates for visual network exploration
Information Visualization
ggplot2: Elegant Graphics for Data Analysis
ggplot2: Elegant Graphics for Data Analysis
Network-theoretic classification of parallel computation patterns
International Journal of High Performance Computing Applications
Hi-index | 0.00 |
A hive plot is a network layout algorithm that uses a parallel coordinate plot in which axes are radially arranged and node position is based on structural properties of that node [8]. We apply hive plots to message-passing communication networks formed by different high performance computing applications at Lawrence Berkeley National Laboratory. Hive plots have advantages over common network visualization methods, abandoning the popular "hairballs" of force-directed network layouts and opting instead for a tunable, repeatable, and interpretable view of network-theoretic properties. Small multiples of these hive plots, called hive panels, are analyzed to suggest which properties contribute to accurate classification of application behavior for anomaly detection in high performance computing environments.