Premium
Sclow Plots: Visualizing Empty Space
Author(s) -
Giesen J.,
Kühne L.,
Lucas P.
Publication year - 2017
Publication title -
computer graphics forum
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.578
H-Index - 120
eISSN - 1467-8659
pISSN - 0167-7055
DOI - 10.1111/cgf.13175
Subject(s) - point cloud , scatter plot , outlier , computer science , point (geometry) , field (mathematics) , inference , dimension (graph theory) , distribution (mathematics) , data point , point distribution model , data mining , pattern recognition (psychology) , algorithm , artificial intelligence , mathematics , geometry , machine learning , pure mathematics , mathematical analysis
Scatter plots are mostly used for correlation analysis, but are also a useful tool for understanding the distribution of high‐dimensional point cloud data. An important characteristic of such distributions are clusters, and scatter plots have been used successfully to identify clusters in data. Another characteristic of point cloud data that has received less attention so far are regions that contain no or only very few data points. We show that augmenting scatter plots by projections of flow lines along the gradient vector field of the distance function to the point cloud reveals such empty regions or voids. The augmented scatter plots, that we call sclow plots, enable a much better understanding of the geometry underlying the point cloud than traditional scatter plots, and by that support tasks like dimension inference, detecting outliers, or identifying data points at the interface between clusters. We demonstrate the feasibility of our approach on synthetic and real world data sets.