Research On Two-dimentional Time Series Data Visualization

A time series is a series of data points indexed in time order,by representing the changes over time,it has been used in statistics,signal processing,mathematical finance,weather forecasting etc.One intuitive way for time series analysis is visualization.Data visualization procedure includes data selection,chart type selection and adjustment of visual parameters.Traditional methods require user interaction which makes it cumbersome.In this paper,by focusing on the task,visualizing main trends in time series,we propose new methods that automatically choose the right visualization type between line graph and scatter plot,and select the proper aspect ratio visual parameter for line graph.Assuming that the main information in a time series is its overall trend,we propose an algorithm that automatically picks the visualization type(line graph or scatter plot)that reveals this trend best.This is achieved by measuring the visual consistency between the trend curve represented by a LOESS fit and the trend described by a scatter plot or a line graph and the consistency score is measured by matching the trend density field with the density fields of the two visualizations using the Earth Mover's Distance(EMD).After choosing the right visualization,the next step is to decide its corresponding visual parameters,here we focus on aspect ratio.In order to choose proper aspect ratios for line graphs,we propose a general framework based on evaluating line integrals for the plotted curves.It provides a unified view for understanding existing methods.Based on these observations,we propose our method of two-scale banking,a technique that combines allow us perceive different patterns within the data.