The visual analysis of flow pattern for public bicycle system

Abstract

Public Bicycle System(PBS) is an increasingly popular mode of public transit, with the advantage of pollution-free and flexibility. In this paper, we present an interactive visual analytic system for exploring complex flows generated by PBS. Four inter-linked visualization views are designed to illustrate multiple perspectives of data, such as the spatial-temporal changes, the relationships and differences between flow OD pairs and the multi-dimensional factors(weather condition, calendar events) influencing on the rental numbers. A new presentation “Parallel coordinates with line and set” combined with flexible interaction schemes is proposed to support the exploration of multivariate association. We exemplify our approach with a real citywide PBS dataset. The results of case study demonstrate that our system is helpful for visually classifying stations with different flow patterns, speculating in-depth reasons, as well as investigating abnormal behaviors, helping decision makers to gain a better understanding of the large dataset.