Web页面中的大规模数据折线图优化方法研究

发布时间：2019-05-08 07:58

【摘要】：在大数据时代,数据可视化在信息挖掘过程中具有极其重要的作用,其中对大规模数据的处理能力是最基础的能力。目前,当数据规模过大时,web页面中的可视化工具都不同程度地出现了一定的性能瓶颈,极大地影响了工作效率。但目前的前端可视化优化手段还比较局限,例如减小文件体积、使用缓存、减少页面节点数量等传统的技术优化手段,对技术优化后带来的信息缺失问题仍没有明确的解决方案。本文选择具有典型代表性和普遍意义的可视化展示工具折线图作为研究对象,提出了在大规模数据场景下的综合优化方案,达到了从性能表现和设计表达两个方面都有了较大提升的效果。方案以D3为绘制基础,采用svg矢量图作为底层机制,保留了现代折线图较强的交互性。借鉴"瓦片地图",采用分层聚合的方式减小浏览器实际渲染的数据规模,从而减轻浏览器的渲染压力。对由于聚合而损失的部分数据特征,在满足条件的情况下,分别使用特定的颜色、灰度和形状来进行表达,并用连线的宽度来暗示数据的缩放级别,表示该折线对应的数据量。将优化方案应用在一个十万级别数据源的折线图实际优化过程中,其绘制结果证明了无论在性能优化还是表达优化方面,本文提出的综合优化方案都起到了有效的作用。
[Abstract]:In big data's era, data visualization plays a very important role in the process of information mining, among which the ability to process large-scale data is the most basic. At present, when the data size is too large, the visualization tools in web pages have some performance bottlenecks in varying degrees, which has a great impact on the efficiency of work. However, the current front-end visualization optimization methods are still relatively limited, such as reducing file size, using cache, reducing the number of page nodes and other traditional technical optimization methods. There is still no clear solution to the lack of information caused by technology optimization. In this paper, a typical representative and universal visual display tool line diagram is selected as the research object, and a comprehensive optimization scheme in large-scale data scenarios is proposed. Achieved from the performance and design expression two aspects have greatly improved the effect. The scheme is based on D3 and the svg vector graph is used as the underlying mechanism, which retains the strong interaction of the modern line graph. Referring to "tile map", the hierarchical aggregation method is adopted to reduce the actual rendering data size of the browser, so as to reduce the rendering pressure of the browser. For some data features lost due to aggregation, the specific color, grayscale and shape are used to express them respectively, and the width of the line is used to indicate the scaling level of the data, which represents the corresponding data amount of the broken line under the condition that the conditions are satisfied. The optimization scheme is applied to the actual optimization process of a 100-thousand-level data source line chart. The results show that the comprehensive optimization scheme proposed in this paper plays an effective role in both performance optimization and expression optimization.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP393.0

【相似文献】