Skip to main content

Causal Graphs for Python

Project description

Code style: black CircleCI unit-tests Checked with mypy codecov

PyWhy-Graphs

pywhy-graphs is a Python graph library that extends networkx with the notion of a MixedEdgeGraph to implement a light-weight API for causal graphical structures that contain mixed-edges and contain causal graph traversal algorithms.

Note: The API is subject to change without deprecation cycles due to the current work-in-progress MixedEdgeGraph class in networkx. For more information, follow the PR at https://github.com/networkx/networkx/pull/5947

Why?

Representation of causal graphical models in Python are severely lacking.

PyWhy-Graphs implements a graphical API layer for ADMG, CPDAG and PAG. For causal DAGs, we recommend using the networkx.DiGraph class and ensuring acylicity via networkx.is_directed_acyclic_graph function.

Existing packages that aim to represent causal graphs either break from the networkX API, or only implement a subset of the relevant causal graphs. By keeping in-line with the robust NetworkX API, we aim to ensure a consistent user experience and a gentle introduction to causal graphical models. A MixedEdgeGraph instance is a composition of networkx graphs and has a similar API, with the additional notion of an "edge type", which specifies what edge type subgraph any function should operate over. For example:

# adds a directed edge from x to y
G.add_edge('x', 'y', edge_type='directed')

# adds a bidirected edge from x to y
G.add_edge('x', 'y', edge_type='bidirected')

Moreover, sampling from causal models is non-trivial, but a requirement for benchmarking many causal algorithms in discovery, ID, estimation and more. We aim to provide simulation modules that are easily connected with causal graphs to provide a simple robust API for modeling causal graphs and then simulating data.

Documentation

See the development version documentation.

Or see stable version documentation

Installation

Installation is best done via pip or conda. For developers, they can also install from source using pip. See installation page for full details.

Dependencies

We aim to provide a very light-weight dependency structure. Minimally, pywhy-graphs requires:

* Python (>=3.8)
* numpy
* scipy
* networkx

Additional functionality may be required when running unit-tests and documentation.

User Installation

If you already have a working installation of numpy, scipy and networkx, the easiest way to install pywhy-graphs is using pip:

# doesn't work until we make an official release :p
pip install -U pywhy-graphs

To install the package from github, clone the repository and then cd into the directory. You can then use poetry to install:

poetry install

# for vizualizing graph functionality
poetry install --extras viz

# if you would like an editable install of dodiscover for dev purposes
pip install -e .

pip install https://api.github.com/repos/py-why/pywhy-graphs/zipball/main

Contributing

Pywhy-Graphs is always looking for new contributors to help make the package better, whether it is algorithms, documentation, examples of graph usage, and more! Contributing to Pywhy-Graphs will be rewarding because you will contribute to a much needed package for causal inference.

See our contributing guide for more details.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

pywhy_graphs-0.1.0.tar.gz (106.4 kB view details)

Uploaded Source

Built Distribution

pywhy_graphs-0.1.0-py3-none-any.whl (138.8 kB view details)

Uploaded Python 3

File details

Details for the file pywhy_graphs-0.1.0.tar.gz.

File metadata

  • Download URL: pywhy_graphs-0.1.0.tar.gz
  • Upload date:
  • Size: 106.4 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.9.16

File hashes

Hashes for pywhy_graphs-0.1.0.tar.gz
Algorithm Hash digest
SHA256 175107f9dce19aba21ed962bc0cf5ebfec55528dee790d0cf89be1e905c5cb2c
MD5 a76f10e1965340046ca9db1c195af5b2
BLAKE2b-256 232858a207ec5df8fc8515cffc49c9512a0803e8cbd1703888c76221133a7439

See more details on using hashes here.

Provenance

File details

Details for the file pywhy_graphs-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: pywhy_graphs-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 138.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.9.16

File hashes

Hashes for pywhy_graphs-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 a7957bed5dbce6619712cad174b922b9f793de2e641d2aa6d7b71a780302620c
MD5 5296ad01deb12a650ca71e1949bdba96
BLAKE2b-256 a07861bb6bc594d616b98220efad9f25231d7915d562e12d431d215cef4fecfd

See more details on using hashes here.

Provenance

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page