Compute the statistical impact of features given a scikit-learn estimator
Project description
This package let’s you compute the statistical impact of features given a scikit-learn estimator. The computation is based on the mean variation of the difference between quantile and original predictions. The impact is reliable for regressors and binary classifiers.
Currently, all features must consist of pure-numerical, non-categorical values.
featureimpact is being developed by Christian Blume. Contact Christian at chr.blume@gmail.com for any questions or comments.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
featureimpact-1.0.3.tar.gz
(4.9 kB
view hashes)