Variable Importance Matching

Code for the Variable Importance Matching paper.

Usage

This code can be used to run the variable important matching (VIM) method and reproduce the results in the paper. To use, follow the steps below:

1. Clone this repo.
2. Create a conda environment using the environment.yml and activate it.
3. Install remaining dependencies in the requirements.txt file using pip.
4. Add the variable_imp_matching/ folder to your PYTHONPATH.
5. See further details in the Experiments/README.md on which scripts to run to reproduce the results in the paper.

Acknowledgements

Variable importance matching is a product of Duke University's Almost Matching Exactly Lab.

We acknowledge funding from the National Science Foundation and Amazon under grant NSF IIS-2147061.

Paper Citation

Lanners Q, Parikh H, Volfovsky A, Rudin C, Page D. From Feature Importance to Distance Metric: An Almost Exact Matching Approach for Causal Inference. arXiv preprint arXiv:2302.11715. 2023 Feb 23.