Dynamic clustering-based mass time sequence curve visualization method

• Main Authors: XIE RENGAN, WEN JING, WANG FEI, CHEN XINGLEI, ZHANG SHUJUN, HUANG YANHAO, LI WENCHEN, FENG HUA, ZHENG WENTING, SHI XINGHUA
• Format: Patent
• Language: Chinese; English
• Published: 31.05.2022
• Subjects: CALCULATING; COMPUTING; COUNTING; ELECTRIC DIGITAL DATA PROCESSING; PHYSICS

The invention discloses a method for visualizing a large number of time sequence curves based on dynamic clustering. The method comprises the following steps: firstly, reading a set of a large number of time sequence curves; carrying out tree-shaped hierarchical subdivision on the curve set according to a time value range to form an m-way tree, defining and calculating k-dimensional feature vectors and comprehensive feature values F of all curves and curve sub-segments, carrying out specific sorting according to two-dimensional spatial distribution, and storing the k-dimensional feature vectors and the comprehensive feature values F on nodes; initial clustering calculation is carried out on the time sequence curve set based on the feature vectors, and the whole curve is divided into specified clusters; when a user interacts to carry out operations such as amplification, reduction and translation on a rectangular range of a two-dimensional observation viewport of a curve set, a curve segment subset passing thr