How Different is Test Case Prioritization for Open and Closed Source Projects

Reference to This Paper

X. Ling, R. Agrawal, and T. Menzies, “How different is test case prioritization for open and closed source projects,” IEEE Transactions on Software Engineering, 2021.

RQ Results of Prioritization Schemes

• Clone the repository and install the required packages list in requirements.txt.
• To reproduce our results for RQ1 and RQ2, run rqs.py in command line by typing

  $ python3 rqs.py [data set name]
  

  For example:

  $ python3 rqs.py TCT_diaspora@diaspora.csv
  

  Simulation results will store in result.txt file. This file is necessary for Scott-Knott analysis.
• Run Scott-Knott analysis in command line by typing

  $ type result.txt| python2 sk_stats.py --text 30 --latex False
  

  or with Latex text

  $ type result.txt| python2 sk_stats.py --text 30 --latex True
  

  Scott-Knott analysis will show in your terminal.
• Examples of Scott-Knott analysis and run time are showed in result folder and runtime folder.
• To reproduce our results for table 16 in the discussion section, run Discussion1.py in command line by typing

  $ python3 Discussion1.py
  

  The code contains all the projects in an input list called "projects" in the code. If you want to add your own projects, please put the .csv files into the /data folder and put the file name in the "projects" list. (We cannot share our proprietary data, so this project is excluded in this repo)
• To reproduce our results for figure 3 in the discussion section, run Discussion2.py in command line by typing

  $ python3 Discussion2.py
  

  The code contains all the projects and their names in two input lists called "projects" and "project_name" in the code. If you want to add your own projects, please put the .csv files into the /data folder and put the file name in the "projects" and "project_name" list. (We cannot share our proprietary data, so this project is excluded in this repo)