API reference¶
These pages are generated directly from the ml4 source docstrings via
mkdocstrings (static analysis, no import at
build time). The docstrings are the single source of truth.
In short
- Each page documents one public entry point, generated from its docstrings.
- Start with AutoSupervised and FittedPipeline for supervised work; the table lists the other workloads.
- Every configuration field, with its type and default, is in Configuration parameters.
- Terms are defined in the Glossary.
Every config parameter
For the exhaustive catalogue of all configuration fields (type, default, description), see Configuration parameters, generated from the live config schema.
Entry points¶
| Page | Object |
|---|---|
| AutoSupervised | parleml_ml4.AutoSupervised |
| FittedPipeline | parleml_ml4.FittedPipeline |
| Ml4Config & DataConfig | parleml_ml4.Ml4Config, parleml_ml4.DataConfig |
| Presets | parleml_ml4.config.presets |
| Configuration parameters | every field of every config |
| Split policy | parleml_ml4.contracts.split_policy.SplitPolicy |
| Signal check | parleml_ml4.foundation_signal.foundation_signal_check |
| Forecaster | parleml_ml4.forecasting.Forecaster |
| AnomalyDetector | parleml_ml4.unsupervised.AnomalyDetector |
| ClusteringDetector | parleml_ml4.unsupervised.ClusteringDetector |
| UnsupervisedPipeline | parleml_ml4.unsupervised.pipeline.UnsupervisedPipeline |
| Unsupervised results | parleml_ml4.unsupervised.result (Anomaly/Clustering) |
| Recommender | parleml_ml4.recommendation.Recommender |
| LTRRanker | parleml_ml4.ltr.LTRRanker |
| AssociationMiner | parleml_ml4.association.AssociationMiner |
Next: AutoSupervised · Glossary