The paper of CFG2AT: Control Flow Graph and Graph Attention Network based Software Defect Prediction is published in the Journal of IEEE Transactions on Reliability, authored by Haiyang Liu, Zhiqiang Li, Hongyu Zhang, Xiao-Yuan Jing, and Jinhui Liu.

We propose a novel SDP approach called Control Flow Graph and Graph Attention Network based Software Defect Prediction (CFG2AT). CFG2AT is designed to identify software defects automatically and contains a graph-structured attention unit to capture control flow information effectively.

• data : The datasets of CFG2AT for Java and Python languages. It contains graph dir with control flow graph information dir and tradition dir with original Promise and JIRA datasets.
• .py : The scripts of CFG2AT.
• .yaml: The environment configurations.
• papermaterial: Some materials related to the paper.

Install required packages:

conda env create -f cfg2at.yaml
conda activate cfg2at

usage: mainforjava.py [-h] [--runTimes RUNTIMES] [--layers LAYERS] [--hiddens HIDDENS] [--epochs EPOCHS] [--numHeads NUMHEADS]
                      {ant,activemq,lucene,jruby,hbase,hive} trainVersion {ant,activemq,lucene,jruby,hbase,hive} testVersion

usage: mainforpy.py [-h] [--runTimes RUNTIMES] [--layers LAYERS] [--hiddens HIDDENS] [--epochs EPOCHS] [--numHeads NUMHEADS]
                    {pandas} trainVersion {pandas} testVersion

If perform WPDP for Java projects, you run:

python mainforjava.py ant 1.5 ant 1.6

If perform CPDP for Java projects, you run:

python mainforjava.py hbase 0.95.0 activemq 5.3.0

If perform WPDP for Python projects, you run

python mainforpy.py pandas 2.2.0 pandas 2.2.1