CausalEGM is a general causal inference framework for estimating causal effects by encoding generative modeling, which can be applied in both discrete and continuous treatment settings.
CausalEGM simultaneously decouples the dependencies of confounders on both treatment and outcome and maps the confounders to the low-dimensional latent space. By conditioning on the low-dimensional latent features, CausalEGM can estimate the causal effect for each individual or the average causal effect within a population.
CausalEGM was originally developed with Python and TensorFlow. Now both Python and R package for CausalEGM are available! Besides, we provide a console program to run CausalEGM directly without running any script. For more information, checkout the Document.
Note that a GPU is recommended for accelerating the model training. However, GPU is not a must, CausalEGM can be installed on any personal computer (e.g, Macbook) or computational cluster with CPU only.
-
Estimate average treatment effect (ATE).
-
Estimate individual treatment effect (ITE).
-
Estiamte average dose response function (ADRF).
-
Estimate conditional average treatment effect (CATE).
-
Built-in simulation and semi-simulation datasets.
Checkout application examples in the Python Tutorial and R Tutorial.
-
Mar/2023: CausalEGM is available in CRAN as a stand-alone R package.
-
Feb/2023: Version 0.2.6 of CausalEGM is released on Anaconda.
-
Dec/2022: Preprint paper of CausalEGM is out on arXiv.
-
Aug/2022: Version 0.1.0 of CausalEGM is released on PyPI.
Create a CausalEGM/data folder and uncompress the dataset in the CausalEGM/data folder.
-
Twin dataset. Google Drive download link.
-
ACIC 2018 datasets. Google Drive download link.
If you find CausalEGM useful for your work, please consider citing our paper:
Qiao Liu, Zhongren Chen, Wing Hung Wong. CausalEGM: a general causal inference framework by encoding generative modeling[J]. arXiv preprint arXiv:2212.05925, 2022.
Found a bug or would like to see a feature implemented? Feel free to submit an issue.
Have a question or would like to start a new discussion? You can also always send us an e-mail.
Your help to improve CausalEGM is highly appreciated! For further information visit https://causalegm.readthedocs.io/.