nannyml.drift package

Subpackages

• nannyml.drift.multivariate package
  • Subpackages
    • nannyml.drift.multivariate.data_reconstruction package
      • Submodules
      • Module contents
    • nannyml.drift.multivariate.domain_classifier package
      • Submodules
      • Module contents
  • Module contents
• nannyml.drift.univariate package
  • Submodules
    • nannyml.drift.univariate.calculator module
    • nannyml.drift.univariate.methods module
    • nannyml.drift.univariate.result module
  • Module contents

Submodules

• nannyml.drift.ranker module

Module contents

NannyML drift detection module.

This module contains ways to detect both univariate (within a single continuous or categorical column) and multivariate (across multiple columns) drift.

The univariate drift detection methods include:

• Kolmogorov-Smirnov statistic (continuous)

• Wasserstein distance (continuous)

• Chi-squared statistic (categorical)

• L-infinity distance (categorical)

• Jensen-Shannon distance

• Hellinger distance

The multivariate drift detection methods include:

• Data reconstruction error: detects drift by performing dimensionality reduction on the model inputs using PCA and then applying the inverse transformation on the latent (reduced) space.

• Domain Classifer: detects drift by looking at how performance a domain classifier is at distinguising between the reference and the chunk datasets.