Module classification 

Supervised classification algorithms.

Each fitted model reports its scores through the shared ClassificationResult parameter struct.

The first classifier is deterministic, closed-form Naive Bayes in two variants — Gaussian (continuous features) and Categorical (discrete features) — living in naive_bayes. Both reproduce scikit-learn’s sklearn.naive_bayes.GaussianNB / CategoricalNB semantics exactly: all scoring is done in log space (no underflow), priors are the training class frequencies, and argmax ties break toward the lower class label. Each fitted model can emit a populated ClassificationResult via its classification_result method, computing accuracy plus macro-averaged precision / recall / F1 from predictions against the true labels.

This module also houses the input-validation, log-space, and metric helpers shared by every classifier in the family; they are module-private and reached from the submodules.

Re-exports

pub use types::*;

Modules

naive_bayes

Naive Bayes classifiers, for the shared classification parameter structs.

types

Plain-data parameter structs for the library’s statistical constructs.