This repository contains the data and scripts of the paper:
E-APR: Mapping the Effectiveness of Automated Program Repair. Aldeida Aleti and Matias Martinez.
arXiv:2002.03968 (2020)
If you use the data and/or scripts, please cite the paper.
@techreport{eapr2020,
title = {E-APR: Mapping the Effectiveness of Automated Program Repair},
year = {2020},
author = {Aldeida Aleti and Matias Martinez},
url = {http://arxiv.org/pdf/2002.03968},
number = {2002.03968},
institution = {arXiv},
}
We consider 3 sets of features: 1) object-oriented features from ckjm_ext, 2) java features from jct, and 3) code features from Coming.
We present 3 sets of plots:
Footprints all repaired bugs from RepairThemAll: 5 bug benchmarks 11 repair tools
Footprints all correctly repaired bugs from RepairThemAll: 5 bug benchmarks 11 repair tools
Footprints all correctly repaired bugs from Defects4J bug benchmark and 16 repair tools
Aldeida Aleti aldeida.aleti@monash.edu
Matias Martinez matias.sebastian.martinez@gmail.com