An Exploratory Analysis for Studying Software Evolution: Time-Delayed Correlation Analysis

We propose an exploratory analysis method for software repositories so that you can study software evolution without an explicit goal. The method extends the traditional correlation analysis to consider a time lag between any pair of metrics. Using data collected from three kinds of repositories, the method automatically calculates optimal parameters to deal with a time lag between metrics and returns the highest correlation coefficient for each pair of metrics. Our case study using one hundred metrics data from three OSS projects shows that the method allows us to find unexpected relationships in some pairs of metrics if we only use the traditional correlation analysis and that using such relationships can be used to build an analysis goal to understand software evolution.