Explaining Bugs with LLMs

Machine Learning for Software Engineering Seminar @HPI

Setup 🛠

Python and a package manager like pip and/or conda need to be installed on your system.

1. Open your terminal and clone the repository:

   git clone https://github.com/valeriatisch/ml4testing.git

   To update the repo with remote changes, run:

   git pull

2. We recommend using a virtual environment. You can create and activate a new environment with conda or another package manager of your preference.

    conda create -n ml4testing pip python=3.8
    conda activate ml4testing

   You will be asked to proceed; type y and enter. This might take a while.

   To deactivate the environment, run:

   conda deactivate

   To remove the environment, run:

   conda remove -n ml4testing --all

   Hint: You can use direnv to automatically (de)activate the environment when you enter the directory. Install direnv and run direnv allow . inside the repo. That is what the .envrc file is for.

3. Install the dependencies:

   pip install -r requirements.txt

4. Create a .env file in the root directory of the project and add the following variables:

   GITHUB_TOKEN=<your github token>
   REPO=<your username>/<your reponame>
   OPENAI_API_KEY=<your openai api key>

   Find an example in .env.example.
   Note:

   • The GitHub Token needs to have permissions to read and write issues on this repository.
   • The OpenAI API Key can be generated here. You need to have a payment method set.

5. You can run the app with:

   python main.py

Development 🚀

Setting Up Pre-Commit in Your Local Development Environment

To keep the code clean and consistent, we use pre-commit as a linter and black as a code formatter.

The configuration is stored in .pre-commit-config.yaml.

Inside the repo, run:

pre-commit install

Now, pre-commit will run automatically on git commit.

If you want to run it manually on all files, use pre-commit run --all-files or pre-commit run <file> for a specific file. This works only on files that are added.

Run black . to format all files automatically or black <file> for a specific file.