Counterfactual fairness

py-why / dowhy

DoWhy is a Python library for causal inference that supports explicit modeling and testing of causal assumptions. DoWhy is based on a unified language for causal inference, combining causal graphical models and potential outcomes frameworks.

https://www.pywhy.org/dowhy

MIT License

6.88k stars 916 forks source link

Counterfactual fairness #1203

Open kmhj13 opened 3 weeks ago

kmhj13 commented 3 weeks ago

Estimates Counterfactual Fairness following Kusner et al. (2018) using DoWhy Reference - https://arxiv.org/pdf/1703.06856.pdf

amit-sharma commented 2 weeks ago

thanks for adding this @kmhj13. I've enabled the automated checks. Will review shortly

bloebp commented 2 weeks ago

Thanks a lot! Looks good to me, just two points:

Let's add it to the https://www.pywhy.org/dowhy/v0.11.1/example_notebooks/nb_index.html#real-world-inspired-examples overview as well.
The notebook is currently not compiling due to the seaborn dependency, which are not required for DoWhy. Can you remove it and use plain Matplotlib?

kmhj13 commented 2 weeks ago

Thanks @bloebp , @amit-sharma ! Have made the changes requested above.

bloebp commented 1 week ago

The path to the datasets seems to be off. It is not referencing the datasets folder.

bloebp commented 1 week ago

Thanks a lot for the changes!

Looks good to me now. @amit-sharma any other concerns?