Causaltestdata is a Python toolkit to generate pseudo dataset for causal inference and causal discovery.
This toolkit assumes all causality is linear in the potential model (i.e., using Structural equation modelling). The input model can be specified as a DAG in networkx graph format, and the generated data is in pandas DataFrame format.
- Source: https://github.com/cpflat/causaltestdata
- Bug Reports: https://github.com/cpflat/causaltestdata/issues
- Author: Satoru Kobayashi
- License: BSD-3-Clause
See notebooks in tutorial directory.