pandrs 0.3.0

A high-performance DataFrame library for Rust, providing pandas-like API with advanced features including SIMD optimization, parallel processing, and distributed computing capabilities
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
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251

//! Regression model evaluation metrics

use crate::error::{Error, Result};
use crate::optimized::OptimizedDataFrame;
use crate::series::Series;
use std::cmp::Ordering;

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_mean_squared_error() {
        let y_true = vec![3.0, 5.0, 2.5, 7.0, 10.0];
        let y_pred = vec![2.8, 4.8, 2.7, 7.2, 9.8];

        let mse = mean_squared_error(&y_true, &y_pred).expect("operation should succeed");

        // Actual calculation: (0.2)² + (0.2)² + (0.2)² + (0.2)² + (0.2)² = 0.2
        assert!((mse - 0.2 / 5.0).abs() < 1e-6); // 0.04 is the correct expected value, compared to six decimal places
    }

    #[test]
    fn test_r2_score() {
        let y_true = vec![3.0, 5.0, 2.5, 7.0, 10.0];
        let y_pred = vec![2.8, 4.8, 2.7, 7.2, 9.8];

        let r2 = r2_score(&y_true, &y_pred).expect("operation should succeed");
        assert!(r2 > 0.99); // R² score should be greater than 0.99
    }

    #[test]
    fn test_empty_input() {
        let empty: Vec<f64> = vec![];

        let mse_result = mean_squared_error(&empty, &empty);
        assert!(mse_result.is_err());

        let r2_result = r2_score(&empty, &empty);
        assert!(r2_result.is_err());
    }

    #[test]
    fn test_different_length() {
        let y_true = vec![1.0, 2.0, 3.0];
        let y_pred = vec![1.0, 2.0];

        let mse_result = mean_squared_error(&y_true, &y_pred);
        assert!(mse_result.is_err());

        let r2_result = r2_score(&y_true, &y_pred);
        assert!(r2_result.is_err());
    }
}

/// Calculate Mean Squared Error (MSE)
///
/// # Arguments
/// * `y_true` - True values
/// * `y_pred` - Predicted values
///
/// # Returns
/// * `Result<f64>` - Mean Squared Error
pub fn mean_squared_error(y_true: &[f64], y_pred: &[f64]) -> Result<f64> {
    if y_true.len() != y_pred.len() {
        return Err(Error::DimensionMismatch(format!(
            "Length mismatch between true and predicted values: {} vs {}",
            y_true.len(),
            y_pred.len()
        )));
    }

    if y_true.is_empty() {
        return Err(Error::InvalidOperation(
            "Cannot calculate with empty data".to_string(),
        ));
    }

    let sum_squared_error = y_true
        .iter()
        .zip(y_pred.iter())
        .map(|(&true_val, &pred_val)| {
            let error = true_val - pred_val;
            error * error
        })
        .sum::<f64>();

    Ok(sum_squared_error / y_true.len() as f64)
}

/// Calculate Mean Absolute Error (MAE)
///
/// # Arguments
/// * `y_true` - True values
/// * `y_pred` - Predicted values
///
/// # Returns
/// * `Result<f64>` - Mean Absolute Error
pub fn mean_absolute_error(y_true: &[f64], y_pred: &[f64]) -> Result<f64> {
    if y_true.len() != y_pred.len() {
        return Err(Error::DimensionMismatch(format!(
            "Length mismatch between true and predicted values: {} vs {}",
            y_true.len(),
            y_pred.len()
        )));
    }

    if y_true.is_empty() {
        return Err(Error::InvalidOperation(
            "Cannot calculate with empty data".to_string(),
        ));
    }

    let sum_absolute_error = y_true
        .iter()
        .zip(y_pred.iter())
        .map(|(&true_val, &pred_val)| (true_val - pred_val).abs())
        .sum::<f64>();

    Ok(sum_absolute_error / y_true.len() as f64)
}

/// Calculate Root Mean Squared Error (RMSE)
///
/// # Arguments
/// * `y_true` - True values
/// * `y_pred` - Predicted values
///
/// # Returns
/// * `Result<f64>` - Root Mean Squared Error
pub fn root_mean_squared_error(y_true: &[f64], y_pred: &[f64]) -> Result<f64> {
    let mse = mean_squared_error(y_true, y_pred)?;
    Ok(mse.sqrt())
}

/// Calculate R² score (coefficient of determination)
///
/// # Arguments
/// * `y_true` - True values
/// * `y_pred` - Predicted values
///
/// # Returns
/// * `Result<f64>` - R² score (1 is best, can be negative if the model is worse than a constant model)
pub fn r2_score(y_true: &[f64], y_pred: &[f64]) -> Result<f64> {
    if y_true.len() != y_pred.len() {
        return Err(Error::DimensionMismatch(format!(
            "Length mismatch between true and predicted values: {} vs {}",
            y_true.len(),
            y_pred.len()
        )));
    }

    if y_true.is_empty() {
        return Err(Error::InvalidOperation(
            "Cannot calculate with empty data".to_string(),
        ));
    }

    // Calculate the mean of true values
    let y_mean = y_true.iter().sum::<f64>() / y_true.len() as f64;

    // Calculate total sum of squares (variance of true values)
    let ss_tot = y_true
        .iter()
        .map(|&true_val| {
            let diff = true_val - y_mean;
            diff * diff
        })
        .sum::<f64>();

    // Calculate residual sum of squares
    let ss_res = y_true
        .iter()
        .zip(y_pred.iter())
        .map(|(&true_val, &pred_val)| {
            let error = true_val - pred_val;
            error * error
        })
        .sum::<f64>();

    // Handle edge cases
    if ss_tot == 0.0 {
        if ss_res == 0.0 {
            Ok(1.0) // Perfect prediction
        } else {
            Ok(0.0) // Constant prediction with error
        }
    } else {
        Ok(1.0 - (ss_res / ss_tot))
    }
}

/// Calculate Explained Variance Score
///
/// # Arguments
/// * `y_true` - True values
/// * `y_pred` - Predicted values
///
/// # Returns
/// * `Result<f64>` - Explained Variance Score (1 is best)
pub fn explained_variance_score(y_true: &[f64], y_pred: &[f64]) -> Result<f64> {
    if y_true.len() != y_pred.len() {
        return Err(Error::DimensionMismatch(format!(
            "Length mismatch between true and predicted values: {} vs {}",
            y_true.len(),
            y_pred.len()
        )));
    }

    if y_true.is_empty() {
        return Err(Error::InvalidOperation(
            "Cannot calculate with empty data".to_string(),
        ));
    }

    // Calculate the mean of true and predicted values
    let y_true_mean = y_true.iter().sum::<f64>() / y_true.len() as f64;
    let y_pred_mean = y_pred.iter().sum::<f64>() / y_pred.len() as f64;

    // Calculate variance of true values
    let var_y_true = y_true
        .iter()
        .map(|&val| {
            let diff = val - y_true_mean;
            diff * diff
        })
        .sum::<f64>()
        / y_true.len() as f64;

    // Calculate variance of residuals
    let var_residual = y_true
        .iter()
        .zip(y_pred.iter())
        .map(|(&t, &p)| {
            let residual = (t - p) - (y_true_mean - y_pred_mean);
            residual * residual
        })
        .sum::<f64>()
        / y_true.len() as f64;

    // Handle edge cases
    if var_y_true == 0.0 {
        if var_residual == 0.0 {
            Ok(1.0)
        } else {
            Ok(0.0)
        }
    } else {
        Ok(1.0 - (var_residual / var_y_true))
    }
}