scry-learn 0.1.0

Machine learning toolkit in pure Rust
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44

// SPDX-License-Identifier: MIT OR Apache-2.0
//! Data preprocessing transformers.
//!
//! Provides scalers, encoders, dimensionality reduction, and a
//! composable [`Transformer`] trait for building preprocessing pipelines.

mod column_transformer;
mod encoder;
mod imputer;
mod normalizer;
mod one_hot;
mod pca;
mod polynomial;
mod scaler;

pub use column_transformer::ColumnTransformer;
pub use encoder::LabelEncoder;
pub use imputer::{SimpleImputer, Strategy};
pub use normalizer::{Norm, Normalizer};
pub use one_hot::{DropStrategy, OneHotEncoder, UnknownStrategy};
pub use pca::Pca;
pub use polynomial::PolynomialFeatures;
pub use scaler::{MinMaxScaler, RobustScaler, StandardScaler};

use crate::dataset::Dataset;
use crate::error::Result;

/// A data transformer that can be fitted on a dataset and applied to transform it.
pub trait Transformer {
    /// Learn parameters from the training data.
    fn fit(&mut self, data: &Dataset) -> Result<()>;

    /// Apply the learned transformation to a dataset (in-place).
    fn transform(&self, data: &mut Dataset) -> Result<()>;

    /// Convenience: fit + transform in one call.
    fn fit_transform(&mut self, data: &mut Dataset) -> Result<()> {
        self.fit(data)?;
        self.transform(data)
    }

    /// Reverse the transformation (if invertible).
    fn inverse_transform(&self, data: &mut Dataset) -> Result<()>;
}