Metadata-Version: 2.1
Name: sdcd
Version: 0.1.4
Summary: Stable differentiable causal discovery for interventional data.
License: MIT
Author: Achille Nazaret
Author-email: aon2108@columbia.edu
Requires-Python: >=3.9
Classifier: Development Status :: 4 - Beta
Classifier: Intended Audience :: Science/Research
Classifier: License :: OSI Approved :: MIT License
Classifier: Natural Language :: English
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Classifier: Programming Language :: Python :: 3.12
Provides-Extra: benchmark
Provides-Extra: tests
Requires-Dist: click (>=8.1.7,<9.0.0) ; extra == "benchmark"
Requires-Dist: networkx (>=3.2.1,<4.0.0)
Requires-Dist: numba (>=0.59.1,<0.60.0)
Requires-Dist: pandas (>=1.1.1,<2.0.0)
Requires-Dist: scipy (>=1.11.3,<2.0.0)
Requires-Dist: torch (>=2.1.0,<3.0.0)
Requires-Dist: tqdm (>=4.66.1,<5.0.0)
Requires-Dist: wandb (>=0.15.12,<0.16.0)
Description-Content-Type: text/markdown

# SDCD: Stable Differentiable Causal Discovery

SDCD is a method for inferring causal graphs from labeled interventional data.\
You can read the associated preprint, ["Stable Differentiable Causal Discovery"](https://arxiv.org/abs/2311.10263), on arXiv.

![sdci-cartoon](https://github.com/azizilab/sdcd/assets/14086852/c2b34538-c5e0-48f2-b092-9167c02101a7)

If you find this work useful, please consider citing our work:

```bibtex
@article{nazaret2023stable,
  title={Stable Differentiable Causal Discovery},
  author={Achille Nazaret and Justin Hong and Elham Azizi and David Blei},
  journal={arXiv preprint arXiv:2311.10263},
  year={2023}
}
```

---

## Quick Start

You can install the package via `pip install sdcd`.

For the main implementation of the method, see the [SDCD](sdcd/models/_sdcd.py) class.

For a tutorial on the basic usage of SDCD, see [this notebook](tutorials/SDCD_basic_usage.ipynb).

Code used to generate paper figures can be found in [this folder](paper_experiments/).

