Jupyter notebook toolbox for model interpretability/explainability
ExpyBox is a Jupyter notebook toolbox for model interpretability/explainability. It lets you create interactive Jupyter notebooks to explain your model.
This package is meant to be used inside of Jupyter notebook, other usage makes little to no sense. First you need to import and instantiate the ExpyBox class:
from expybox import ExpyBox expybox = ExpyBox(train_data, predict_function, kernel_globals=globals())
Now you can use the supported interpretability methods like this (for list of supported methods refer to the documentation):
which creates a form:
In this form you can set up explained instance (if it's necessary for the selected method)
and method parameters. After clicking on
Run Interact the method will be executed
and its output will be shown below the form.
You can then change the parameters or the explained instance and press
again which will rerun the method with new parameters.
You can find an example Jupyter notebook in
Because of alibi package ExpyBox requires 64-bit Python 3.7 or higher. It is also recommended to create separate virtual enviroment - you can use Pythons venv.
Otherwise the installation process is the same as for other packages, just use pip:
pip install expybox
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size expybox-1.0.1-py3-none-any.whl (21.2 kB)||File type Wheel||Python version py3||Upload date||Hashes View hashes|
|Filename, size expybox-1.0.1.tar.gz (16.7 kB)||File type Source||Python version None||Upload date||Hashes View hashes|