zuomatthew / causalgraphicalmodels

Causal Graphical Models in Python

CausalGraphicalModels

Introduction

causalgraphicalmodels is a python module for describing and manipulating Causal Graphical Models and Structural Causal Models. Behind the scenes it is a light wrapper around the python graph library networkx, together with some CGM specific tools.

It is currently in a very early stage of development. All feedback is welcome.

Example

For a quick overview of CausalGraphicalModel, see this example notebook.

Install

pip install causalgraphicalmodels

Resources

My understanding of Causality comes mainly from the reading of the follow work:

• Causality, Pearl, 2009, 2nd Editing. (An overview available here)
• A fantastic blog post, If correlation doesn’t imply causation, then what does? from Michael Nielsen
• These lecture notes from Jonas Peters
• The draft of Elements of Causal Inference
• http://mlss.tuebingen.mpg.de/2017/speaker_slides/Causality.pdf

Related Packages

• Causality
• CausalInference
• DoWhy