Skip to content

Latest commit

 

History

History
41 lines (29 loc) · 1.69 KB

README.md

File metadata and controls

41 lines (29 loc) · 1.69 KB
Raw

Bug Detection and Repair

yeet

Bug Detection and Repair Dataset with Exploratory Data Analysis examples and demo applications for Bug Detection and Repair algorithm.

Content

Source code related tasks for machine learning have become important with the large need of software production. In this project our main goal is to create a dataset for bug detection and repair and also provide examples of code repair algorithms.

This repository contains the scripts used for dataset generation. We have also added some exploratory data analysis notebooks for the generated datasets. The bugnet folder contains the script used to generate the BugNet dataset. The repair-pipeline folder contains the demo applications for the models trained only on the Python code from BugNet. The aoc-dataset folder contains the source code used to generate the AoC dataset. The hint folder contains the source code used to generate the hint in natural language of the bugs. The repair folder contains the source code used to evaluate different models on the data that we collected.

To install the dependencies for development create a venv:

python -m venv .venv
source .venv/bin/activate
make install

  1. To run the repair pipeline see repair-pipeline

  2. To generate the BugNet dataset see bugnet

  3. To visualize the AoC dataaset see aoc-dataset

  4. To visualize the results of the hint generation on the BugNet (of AoC) dataset see hint

  5. To visualize the result of the repair generation on the BugNet (or AoC) dataset see repair