Statistical properties analysis of file modification in open-source software repositories
Yongfeng Tian, Huobin Tan, Guangyan Lin · 2018
Mining the evolutionary rules of source code files can be conducted by analyzing the data generated in the development of open source software. In this paper, the development log information of two famous open source projects is collected and the statistical distribution of the number of developers corresponding to class files modification is analyzed by statistical method. As a result, we discover that the statistical distribution of the number of developers corresponding to class files modification fellows approximately an exponential distribution. In addition, we analyze the features of function and structure of two kinds of class files and discover that both class files developed by developers who have too many modification behaviors to their projects and class files modified by too many developers tend to be more complex. The statistical analysis of these two projects may provide new insights into the research on studying the evolution of open source software, choosing appropriate programmers to refactor open source software and allocating task of maintenance for open source software.