datarust 0.4.1

Scikit-Learn Preprocessing in 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
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117

//! Pipeline workflow: chain transformers, test inverse-transform round-trip,
//! pipeline ergonomics, and feature names.
//!
//! Run: `cargo run --example pipeline_workflow`

use datarust::compose::{ColumnTransformer, Remainder, Table};
use datarust::encoder::OneHotEncoder;
use datarust::pipeline::Pipeline;
use datarust::scaler::{MinMaxScaler, StandardScaler};
use datarust::selection::VarianceThreshold;
use datarust::transformer_kind::TransformerKind;
use datarust::CategoricalTransformerKind;
use datarust::FeatureNames;
use datarust::Transformer;
use datarust::{Matrix, StrMatrix};

fn main() -> Result<(), Box<dyn std::error::Error>> {
    // ── 1. Numeric-only pipeline with inverse transform ────────────────
    let x = Matrix::new(vec![
        vec![1.0, 500.0],
        vec![2.0, 300.0],
        vec![3.0, 800.0],
        vec![4.0, 100.0],
    ])?;

    let mut pipeline = Pipeline::new()
        .push(
            "scale",
            TransformerKind::StandardScaler(StandardScaler::new()),
        )
        .push("norm", TransformerKind::MinMaxScaler(MinMaxScaler::new()));

    let transformed = pipeline.fit_transform(&x)?;
    println!("=== Pipeline (StandardScaler → MinMaxScaler) ===");
    println!("Input:  {}×{}", x.nrows(), x.ncols());
    println!("Output: {}×{}", transformed.nrows(), transformed.ncols());
    for i in 0..transformed.nrows() {
        let row: Vec<String> = transformed
            .row(i)
            .iter()
            .map(|v| format!("{:.4}", v))
            .collect();
        println!("  [{}]", row.join(", "));
    }

    // ── 2. Pipeline ergonomics: step inspection and replacement ───────
    println!("=== Pipeline ergonomics ===");
    println!("Step names: {:?}", pipeline.names());

    // Access a step by name
    if let TransformerKind::MinMaxScaler(mm) = pipeline.get_step("norm").unwrap() {
        println!(
            "  norm step: MinMaxScaler(range={:?})",
            mm.feature_range_value()
        );
    }

    // Replace a step
    pipeline.set_step(
        "norm",
        TransformerKind::VarianceThreshold(VarianceThreshold::new(0.01)?),
    );
    println!("  after replacing norm with VarianceThreshold:");
    println!("  Step names: {:?}", pipeline.names());

    // ── 3. Feature names through pipeline ─────────────────────────────
    let names = pipeline.feature_names_out(Some(&["feat0".into(), "feat1".into()]));
    println!("\n=== Feature names ===");
    println!("{:?}", names);

    // ── 4. ColumnTransformer with categorical data ────────────────────
    let numeric = Matrix::new(vec![
        vec![10.0, 1000.0],
        vec![20.0, 2000.0],
        vec![30.0, 3000.0],
    ])?;
    let categorical = StrMatrix::from_column(["X", "Y", "Z"])?;
    let table = Table::new(numeric, categorical)?;

    let mut ct = ColumnTransformer::new()
        .remainder(Remainder::Passthrough)
        .add_numeric(
            "scaled",
            vec![0],
            TransformerKind::StandardScaler(StandardScaler::new()),
        )
        .add_categorical(
            "cat",
            vec![0],
            CategoricalTransformerKind::OneHotEncoder(OneHotEncoder::new()),
        );

    let out = ct.fit_transform(&table)?;
    println!("=== ColumnTransformer (passthrough) ===");
    println!("Output: {} rows × {} cols", out.nrows(), out.ncols());
    for i in 0..out.nrows() {
        let row: Vec<String> = out.row(i).iter().map(|v| format!("{:.4}", v)).collect();
        println!("  [{}]", row.join(", "));
    }

    // Feature names
    let names = ct.feature_names_out(None);
    println!("\nFeature names: {:?}", names);

    // ── 5. Multi-output (table) transform ────────────────────────────
    let output = ct.fit_transform_to_table(&table)?;
    println!("\n=== Multi-output ===");
    println!(
        "Numeric: {}×{} | Categorical: {}×{}",
        output.numeric.nrows(),
        output.numeric.ncols(),
        output.categorical.nrows(),
        output.categorical.ncols(),
    );

    Ok(())
}