Expand description
§ferrolearn
A scikit-learn equivalent for Rust.
This is the top-level re-export crate that provides a unified API for all ferrolearn functionality. Individual sub-crates can also be used directly for finer-grained dependency control.
Modules§
- bayes
- ferrolearn-bayes
- cluster
- ferrolearn-cluster
- core
- ferrolearn-core
- datasets
- ferrolearn-datasets
- decomp
- ferrolearn-decomp
- io
- ferrolearn-io
- linear
- ferrolearn-linear
- metrics
- ferrolearn-metrics
- model_
selection - ferrolearn-model-sel
- neighbors
- ferrolearn-neighbors
- prelude
- Convenience prelude that re-exports the most commonly used traits and types.
- preprocess
- ferrolearn-preprocess
- sparse
- ferrolearn-sparse
- tree
- ferrolearn-tree
Structs§
- Pipeline
- A dynamic-dispatch pipeline that composes transformers and a final estimator.
Enums§
- Ferro
Error - The unified error type for all ferrolearn operations.
Traits§
- Backend
- Trait abstracting core linear algebra operations.
- Dataset
- A trait for types that represent tabular datasets.
- Fit
- Train a model on data, producing a fitted model.
- FitTransform
- Combined fit-and-transform in a single pass.
- Partial
Fit - Incrementally train a model on a batch of data.
- Predict
- Generate predictions from a fitted model.
- Transform
- Transform data (e.g., scaling, encoding).
Type Aliases§
- Default
Backend - The default linear algebra backend.
- Ferro
Result - A convenience type alias for
Result<T, FerroError>.