Complex network theory has stimulated the study of complexities in various kinds of fields. By the algorithm of visibility graph, weekly and daily time series of Shanghai stock index was converted into networks to analyze their characteristics. The overall properties of networks remain similar, while details are different with different levels of details provided by different frequency data. The findings provide evidences for principles of choosing daily time series for the network analysis in previous researches. Also, the community detection of the network provides proper ways for time series trend analysis, and the detected trends are not influenced by series frequencies.
Keywords: Time series, visibility graph, network