Code for our paper, currently considered for acceptance at the 36th Conference on Advanced Information Systems Engineering (CAiSE 2024).
Below you will find instructions on how to use the code.
You can either install dependencies (including the correct python version) via mamba (recommended), or conda (slower). These require an installation of mamba (https://mamba.readthedocs.io/en/latest/installation.html), or conda (https://docs.conda.io/en/latest/miniconda.html) respectively.
Alternatively, you can install the dependencies manually via pip. In this case please refer to the contents of env.yaml to get a list of dependencies and python version.
Below you can find the instructions for installing dependencies with mamba.
Note, that the environment still has to be activated with conda.
mamba env create -f env.yaml
conda activate pet-baselineSimilarly, you can find instructions for installing dependencies with conda. Beware that this may take several (> 5) minutes, depending on your machine!
conda env create -f env.yaml
conda activate pet-baselineNext setup a mysql database named pet_data_augment and a database user
named optuna@localhost (no credentials, socket auth). Alternatively adjust
code in lines 193 - 198 of augmentation_experiments.py, to use e.g. Sqlite, see
https://optuna.readthedocs.io/en/stable/tutorial/20_recipes/001_rdb.html.
The command given below will run the augmentation experiments as defined in our paper.
conda activate pet-baseline
python -m augmentation_experimentsThis will start the hyper-parameter optimization, which takes several days, depending on your machine. In our experiments the data augmentation ran for more than five days.