linreg-core 0.8.1

Lightweight regression library (OLS, Ridge, Lasso, Elastic Net, WLS, LOESS, Polynomial) with 14 diagnostic tests, cross validation, and prediction intervals. Pure Rust - no external math dependencies. WASM, Python, FFI, and Excel XLL bindings.
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
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216

// ============================================================================
// Model Serialization Example
// ============================================================================
//
// This example demonstrates how to save and load trained regression models
// using the ModelSave and ModelLoad traits.
//
// Run with:
//     cargo run --example serialization_example

use linreg_core::core::ols_regression;
use linreg_core::loess::{loess_fit, LoessOptions};
use linreg_core::regularized::{elastic_net_fit, lasso_fit, ridge_fit, ElasticNetOptions, LassoFitOptions, RidgeFitOptions};
use linreg_core::serialization::{ModelLoad, ModelSave};
use linreg_core::weighted_regression::wls_regression;
use linreg_core::linalg::Matrix;

fn main() -> Result<(), Box<dyn std::error::Error>> {
    println!("=== Model Serialization Example ===\n");

    // Sample data
    let y = vec![2.5, 3.7, 4.2, 5.1, 6.3, 7.0, 8.1, 9.2];
    let x1 = vec![1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0];
    let x2 = vec![1.5, 2.1, 3.2, 3.9, 5.1, 6.2, 7.0, 8.1];

    // ========================================================================
    // OLS Regression Example
    // ========================================================================

    println!("--- OLS Regression ---");
    let names = vec!["Intercept".to_string(), "X1".to_string(), "X2".to_string()];
    let ols_result = ols_regression(&y, &[x1.clone(), x2.clone()], &names)?;
    println!("R²: {:.4}", ols_result.r_squared);
    println!("Coefficients: {:?}", ols_result.coefficients);

    // Save with custom name
    let ols_path = "ols_model.json";
    ols_result.save_with_name(ols_path, Some("My OLS Housing Model".to_string()))?;
    println!("Saved to: {}", ols_path);

    // Load back
    let ols_loaded = linreg_core::core::RegressionOutput::load(ols_path)?;
    println!("Loaded model R²: {:.4}", ols_loaded.r_squared);
    println!();

    // ========================================================================
    // Ridge Regression Example
    // ========================================================================

    println!("--- Ridge Regression ---");
    let x_matrix = create_matrix(&y.len(), &[&x1, &x2]);
    let ridge_options = RidgeFitOptions {
        lambda: 1.0,
        standardize: true,
        intercept: true,
        ..Default::default()
    };
    let ridge_result = ridge_fit(&x_matrix, &y, &ridge_options)?;
    println!("Lambda: {}", ridge_result.lambda);
    println!("Intercept: {:.4}", ridge_result.intercept);
    println!("Coefficients: {:?}", ridge_result.coefficients);

    let ridge_path = "ridge_model.json";
    ridge_result.save(ridge_path)?;
    println!("Saved to: {}", ridge_path);

    let ridge_loaded = linreg_core::regularized::ridge::RidgeFit::load(ridge_path)?;
    println!("Loaded model Lambda: {}", ridge_loaded.lambda);
    println!();

    // ========================================================================
    // Lasso Regression Example
    // ========================================================================

    println!("--- Lasso Regression ---");
    let lasso_options = LassoFitOptions {
        lambda: 0.1,
        standardize: true,
        intercept: true,
        ..Default::default()
    };
    let lasso_result = lasso_fit(&x_matrix, &y, &lasso_options)?;
    println!("Lambda: {}", lasso_result.lambda);
    println!("Non-zero coefficients: {}", lasso_result.n_nonzero);
    println!("Coefficients: {:?}", lasso_result.coefficients);

    let lasso_path = "lasso_model.json";
    lasso_result.save(lasso_path)?;
    println!("Saved to: {}", lasso_path);
    println!();

    // ========================================================================
    // Elastic Net Regression Example
    // ========================================================================

    println!("--- Elastic Net Regression ---");
    let enet_options = ElasticNetOptions {
        lambda: 0.1,
        alpha: 0.5,
        standardize: true,
        intercept: true,
        ..Default::default()
    };
    let enet_result = elastic_net_fit(&x_matrix, &y, &enet_options)?;
    println!("Lambda: {}, Alpha: {}", enet_result.lambda, enet_result.alpha);
    println!("Non-zero coefficients: {}", enet_result.n_nonzero);
    println!("Coefficients: {:?}", enet_result.coefficients);

    let enet_path = "elastic_net_model.json";
    enet_result.save(enet_path)?;
    println!("Saved to: {}", enet_path);
    println!();

    // ========================================================================
    // WLS (Weighted Least Squares) Example
    // ========================================================================

    println!("--- Weighted Least Squares (WLS) ---");
    let weights = vec![1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0];
    let wls_result = wls_regression(&y, &[x1.clone(), x2.clone()], &weights)?;
    println!("R²: {:.4}", wls_result.r_squared);
    println!("Coefficients: {:?}", wls_result.coefficients);

    let wls_path = "wls_model.json";
    wls_result.save(wls_path)?;
    println!("Saved to: {}", wls_path);
    println!();

    // ========================================================================
    // LOESS Regression Example
    // ========================================================================

    println!("--- LOESS Regression ---");
    let loess_options = LoessOptions {
        span: 0.75,
        degree: 2,
        robust_iterations: 2,
        n_predictors: 1,
        surface: linreg_core::loess::LoessSurface::Direct,
    };
    let loess_result = loess_fit(&y, &[x1.clone()], &loess_options)?;
    println!("Span: {}, Degree: {}", loess_result.span, loess_result.degree);
    println!("First 3 fitted values: {:?}", &loess_result.fitted[..3]);

    let loess_path = "loess_model.json";
    loess_result.save(loess_path)?;
    println!("Saved to: {}", loess_path);
    println!();

    // ========================================================================
    // Error Handling Examples
    // ========================================================================

    println!("--- Error Handling ---");

    // Try to load with wrong type
    match linreg_core::regularized::ridge::RidgeFit::load(ols_path) {
        Ok(_) => println!("Unexpectedly loaded OLS as Ridge"),
        Err(e) => println!("Expected error loading OLS as Ridge: {}", e),
    }

    // Try to load non-existent file
    match linreg_core::core::RegressionOutput::load("nonexistent.json") {
        Ok(_) => println!("Unexpectedly loaded nonexistent file"),
        Err(e) => println!("Expected error for nonexistent file: {}", e),
    }

    println!();

    // ========================================================================
    // Metadata Inspection
    // ========================================================================

    println!("--- Metadata Inspection ---");
    // The saved files contain metadata that can be inspected
    let content = std::fs::read_to_string(ols_path)?;
    let parsed: serde_json::Value = serde_json::from_str(&content)?;
    if let Some(metadata) = parsed.get("metadata") {
        println!("Model metadata:");
        println!("  Type: {}", metadata.get("model_type").unwrap());
        println!("  Format version: {}", metadata.get("format_version").unwrap());
        println!("  Library version: {}", metadata.get("library_version").unwrap());
        println!("  Created at: {}", metadata.get("created_at").unwrap());
        println!("  Name: {:?}", metadata.get("name").unwrap());
    }

    println!();

    // ========================================================================
    // Cleanup
    // ========================================================================

    println!("--- Cleanup ---");
    for path in &[ols_path, ridge_path, lasso_path, enet_path, wls_path, loess_path] {
        let _ = std::fs::remove_file(path);
        println!("Removed: {}", path);
    }

    println!("\n=== Example Complete ===");

    Ok(())
}

/// Helper function to create a design matrix from predictor columns
fn create_matrix(n: &usize, x_cols: &[&Vec<f64>]) -> Matrix {
    let p = x_cols.len();
    let mut data = vec![1.0; n * (p + 1)]; // First column is intercept

    for (col_idx, col) in x_cols.iter().enumerate() {
        for (row_idx, &val) in col.iter().enumerate() {
            data[row_idx * (p + 1) + col_idx + 1] = val;
        }
    }

    Matrix::new(*n, p + 1, data)
}