smclarify
Amazon Sagemaker Clarify
Bias detection and mitigation for datasets and models.
Installation
To install the package from PIP you can simply do:
pip install smclarify
You can see examples on running the Bias metrics on the notebooks in the examples folder.
Terminology
Facet
A facet is column or feature that will be used to measure bias against. A facet can have value(s) that designates that sample as "sensitive".
Label
The label is a column or feature which is the target for training a machine learning model. The label can have value(s) that designates that sample as having a "positive" outcome.
Bias measure
A bias measure is a function that returns a bias metric.
Bias metric
A bias metric is a numerical value indicating the level of bias detected as determined by a particular bias measure.
Bias report
A collection of bias metrics for a given dataset or a combination of a dataset and model.
Development
It's recommended that you setup a virtualenv.
virtualenv -p(which python3) venv
source venv/bin/activate.fish
pip install -e .[test]
cd src/
../devtool all
For running unit tests, do pytest --pspec. If you are using PyCharm, and cannot see the green run button next to the tests, open Preferences -> Tools -> Python Integrated tools, and set default test runner to pytest.
For Internal contributors, run ../devtool integ_tests after creating virtualenv with the above steps to run the integration tests.
Metadata
Release files for smclarify 0.5
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| smclarify-0.5-py3-none-any.whl | Python 3 | none | any | Details for smclarify-0.5-py3-none-any.whl |
Release files / smclarify-0.5-py3-none-any.whl
| Download URL | smclarify-0.5-py3-none-any.whl |
|---|---|
| Size | 30.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
49ed1d65b4296e8458f00dae156fb4dceb25c4e4752b756a558a5fd2e9379233
|
|
BLAKE2b-256 checksum How to use checksums |
e308db47c6699e2c82a3566bcb9d447a71176634944fb7da0cb5067e6aadc06f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/4.0.2 CPython/3.9.13
|
Security for smclarify 0.5
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.