Mason Lee, Hicks Blánaid, Almeida Jonas S
Division of Cancer Epidemiology and Genetics, National Institutes of Health, Rockville, Maryland, United States of America.
Center for Public Health, Queen's University Belfast, Belfast, United Kingdom.
PLoS One. 2025 May 27;20(5):e0322393. doi: 10.1371/journal.pone.0322393. eCollection 2025.
Spatial cluster analysis is crucial for understanding localized patterns in geospatial data, with wide-ranging applications for scientific discovery and decision-making. However, the dynamic nature of spatial clusters and the diverse range of clustering methods available can make analysis and interpretation challenging. We introduce ClusterRadar, a web-based tool designed to streamline this process by uniquely prioritizing longitudinal analysis and multi-method comparison of spatial clusters. It empowers users to easily perform clustering with multiple methods, directly compare results, and visualize spatiotemporal patterns through a novel design of linked interactive visualizations. ClusterRadar aims to maximize utility to a broad user base by supporting various geospatial formats and executing entirely within the browser to ensure data privacy. ClusterRadar is available at https://episphere.github.io/ClusterRadar.
空间聚类分析对于理解地理空间数据中的局部模式至关重要,在科学发现和决策制定方面有广泛应用。然而,空间聚类的动态性质以及可用聚类方法的多样性会使分析和解释具有挑战性。我们引入了ClusterRadar,这是一个基于网络的工具,旨在通过独特地优先考虑空间聚类的纵向分析和多方法比较来简化这一过程。它使用户能够轻松地用多种方法进行聚类,直接比较结果,并通过一种新颖的链接交互式可视化设计来可视化时空模式。ClusterRadar旨在通过支持各种地理空间格式并完全在浏览器内执行以确保数据隐私,从而为广大用户群体最大化效用。可在https://episphere.github.io/ClusterRadar上获取ClusterRadar。