I. Requirement

Clone the following repos directly into the root folder of this repository.

• Defects4J

Follow the TBar setup from the readme in TBar folder. Make sure to check out all D4J projects.

Ensure you have successfully installed CUDA (version >= 11.4, preferably 11.7) along with the necessary drivers. Additionally, make sure you have installed torch (version 2.0.1).

Next, we have provided an init_env.sh script to simplify the installation of smaller required packages. Execute the following command to run the script:

sh init_env.sh

All LLMs involved will be downloaded automatically during your first run.

II. Replicate results

1. RQ1 Fault localization (FL): The following script prints out fault localization top-score values for entropy and prior FL tools.

   • python3 analysis_notebooks/rq1_fl.py

2. RQ2-3 Patch classification: Run the interactive ipynb notebook.

   • analysis_notebooks/patch_analysis.ipynb. This notebook provides results on patch ranking, patch classification, and generates plots used in the paper.

III. Rerun entropy from scratch

1. The following 3 scripts generate entropy for FL results:

   • python3 fl_scores/process_llmao_prior.py
   • python3 fl_scores/process_sbfl.py
   • python3 fl_scores/process_transfer.py

2. The following scripts generate entropy for patch efficiency results:

   • python3 patches/generate_tbar_patches.py tbar_vanilla
   • python3 patches/generate_tbar_patches.py tbar_testcache
   • python3 patches/generate_tbar_patches.py tbar_testcache_ranked

3. The following 3 scripts generate entropy for patch ranking results.

   • python3 patches/shib_panther_patch_entropy.py
   • python3 patches/tbar_patch_entropy.py