Introduction

This repository includes the experimental implementation of a paper entitled An ensemble-based predictive mutation testing approach that considers impact of unreached mutants (link).

In this paper, we show that the uncovered mutants could reduce the results of predictive mutation testing considerablly. To investigate our hypothesis we perform the following three steps:

Step 1: Replication

We replicate the study of previous predictive mutation testing (link). However, we take uncovered mutants into account when we evaluate the model. The results show that the AUC drops from 0.83 to 0.51.

Step 2: Proposed approach

We proposed an approach heavily based on ensemble techniques. The proposed approach used Random Forest and Gradient Boosting to predict the results of mutant execution.

Structure

In order to use this repository, you need to first download the dataset provided by Mao (link). We would like to thank Mao for making the data publicly available. Our research is impossible without that data.

Afer downloading the data, you should put the unzip thereof in the folder docs/Data . Now, you are able to run the experiments written in the Python programming language. We use jupyter notebooks for running the code.

Citation

Please cite the paper using the following bibtex entry:

@misc{aghamohammadi2020threat,
    title={The Threat to the Validity of Predictive Mutation Testing: The Impact of Uncovered Mutants},
    author={Alireza Aghamohammadi and Seyed-Hassan Mirian-Hosseinabadi},
    year={2020},
    eprint={2005.11532},
    archivePrefix={arXiv},
    primaryClass={cs.SE}
}