API Reference#

The class and function reference for Empulse, organised by module. These pages document the raw specifications: every parameter, attribute and method.

For the reasoning behind a component and worked examples of using it, see the User Guide — the signatures alone rarely convey when you should reach for something.

Metrics

empulse.metrics

Cost matrices, metric strategies, and the prebuilt churn, acquisition and credit scoring metrics.

empulse.metrics
Models

empulse.models

Cost-sensitive classifiers, threshold meta-estimators, and the bias mitigation wrappers.

empulse.models
Samplers

empulse.samplers

Cost-proportionate and fairness-aware resampling, in the imbalanced-learn style.

empulse.samplers
Optimizers

empulse.optimizers

Solvers and learning-rate schedules driving the logit-family and evolutionary models.

empulse.optimizers
Datasets

empulse.datasets

Loaders for the real-world cost-sensitive datasets, each shipping its own cost matrix.

empulse.datasets

Note

Empulse has no top-level re-exports. Import from the submodules — from empulse.models import CSBoostClassifier — rather than from empulse directly.