How to reproduce the results?

Creating conda environment

   conda env create -f environment.yml
   conda activate jit-def-pred
   ipython kernel install --user --name=jit-def-pred

Getting the data

By your own

1. Clone needed repositories:

• Openstack
• Qt

cd jit-defect-prediction/data/repositories
git clone git@github.com:openstack/openstack.git

cd qt
awk '{ system("git clone git@github.com:qt/"$1".git") }' qt_repositories

2. Download files changed by commits:

cd jit-defect-prediction
conda activate jit-def-pred
python src/preparation/download_data.py --project openstack
python src/preparation/download_data.py --project qt

You can also run test to check if the script works correctly:

python src/preparation/download_data.py --project test

Downloading already prepared data

Copy data (around 15GB) from Google Drive. and extract them to jit-defect-prediction/data directory.

Fitting baseline models

How to fit and search for parameters of a baseline model?

cd jit-defect-prediction
conda activate jit-def-pred
python src/models/baseline.py --help

Results can be found in directory data/models

Project information

Google sheet with list of relevant publications, ideas etc.