Efficient Brushing Algorithm for Density-Based Parallel Coordinates Plot
Year: 2026
Type: article
Abstract: Brushing is a fundamental interaction for exploring parallel coordinates plots (PCPs), allowing users to restrict a visualization to records that satisfy selected attribute ranges. For density-based PCPs, however, brushing requires more than identifying the selected records: the density field induced by the selected subset must be reconstructed. On large datasets, exact reconstruction is expensive because every selected segment may contribute to every pixel of the visualization.We study efficient brushing for density-based PCPs and proposeRange-Sampled Density Brushing(RSDB). RSDB localizes the brushed population exactly while reconstructing its density approximately. For each adjacent axis pair, it interprets a brush as a two-dimensional orthogonal range query and uses a range tree to obtain an exact decomposition of the selected subset. It then draws uniform conditional samples directly from this decomposition and estimates the brushed density using only the sampled segments. The sample budget adapts to the brush selectivity and guarantees the maximum absolute error over the entire visualization does not exceed a user-specified tolerance ε. Experiments on four large datasets show that RSDB consistently achieves substantially higher query throughput than exact scan- and index-based baselines across different settings. The empirical approximation errors are also consistent with the prescribed probabilistic guarantee.
Source: IEEE Access
Authors: Yanting Sun, Yuntao Sun, Xuehua Zhang
Institution: Shandong College of Electronic Technology
Language: English
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,075
OpenAlex ID: W7223674041