Skip to content

Simple example of how to use an embedding plus sphering/whitening transform to measure difference in distribution.

License

Notifications You must be signed in to change notification settings

WinVector/TypicalityCoding

Repository files navigation

TypicalityCoding

Materials to support the blog article Detecting Data Differences Using the Sphering Transform.

Simple example notebook of how to use an embedding plus sphering/whitening transform to measure difference in distribution.

Further Examples:

To install environment: conda env create --file=tc_env_def.yaml

Specific versions used saved as: conda list --export > tc_env_list.txt

To use environment: conda activate tc_env

Download data from: https://drive.google.com/drive/folders/1sXtDqNF-D0mypORzfevir_eH9FdG2668?usp=sharing and then unzip

To run tests: python -m pytest tests

About

Simple example of how to use an embedding plus sphering/whitening transform to measure difference in distribution.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published