/dci

Code and figures for the Differential Causal Inference (DCI) algorithm

Primary LanguageJupyter Notebook

This repository contains the code for the Differential Causal Inference (DCI) algorithm.

The arXiv preprint is available at [https://arxiv.org/abs/1802.05631]

Authors: Yuhao Wang, Chandler Squires, Anastasiya Belyaeva, and Caroline Uhler.

This project uses Python 3.6.4 and R 3.5.0

To get started, download, unzip, and run:

  • bash make_venv.sh
  • R -f install_packages.R

To recreate the figures, run

  • source venv/bin/activate
  • bash create_fig1ab_data.sh
  • python3 create_fig1c_data.py
  • bash create_fig2_data.sh
  • ipython3 kernel install --user --name=dci
  • jupyter-notebook

Then run the notebooks figure1ab.ipynb, figure1c.ipynb, and figure2.ipynb