Code Author: Trent Kyono and Boris van Breugel

This repository contains the code used for the "DECAF: Generating Fair Synthetic Data Using Causally-Aware Generative Networks" paper(2021).

pip install -r requirements.txt
pip install .

You can run the tests using

pip install -r requirements_dev.txt
pip install .
pytest -vsx

• decaf/DECAF.py - Synthetic data generator class - DECAF.
• tests/run_example.py - Runs a nonlinear toy DAG example. The dag structure is stored in the dag_seed variable. The edge removal is stored in the bias_dict variable. See example usage in this file.

Base example on toy dag:

$ cd tests
$ python run_example.py

An example to run with a dataset size of 2000 for 300 epochs:

$ python run_example.py --datasize 2000 --epochs 300

@inproceedings{kyono2021decaf,
	title        = {DECAF: Generating Fair Synthetic Data Using Causally-Aware Generative Networks},
	author       = {van Breugel, Boris and Kyono, Trent and Berrevoets, Jeroen and van der Schaar, Mihaela},
	year         = 2021,
	booktitle    = {Conference on Neural Information Processing Systems(NeurIPS) 2021}
}