Function accuracy_score 

pub fn accuracy_score<T, S1, S2, D1, D2>(
    y_true: &ArrayBase<S1, D1>,
    y_pred: &ArrayBase<S2, D2>,
) -> Result<f64>
where
    T: PartialEq + NumCast + Clone,
    S1: Data<Elem = T>,
    S2: Data<Elem = T>,
    D1: Dimension,
    D2: Dimension,

Calculates accuracy score, the fraction of correctly classified samples

Mathematical Formulation

The accuracy score is defined as:

Accuracy = (Number of Correct Predictions) / (Total Number of Predictions)
         = (TP + TN) / (TP + TN + FP + FN)

Where:

• TP = True Positives
• TN = True Negatives
• FP = False Positives
• FN = False Negatives

For multi-class classification, accuracy is simply:

Accuracy = (1/n) * Σ I(ŷᵢ = yᵢ)

Where:

• n = total number of samples
• I(·) = indicator function (1 if condition is true, 0 otherwise)
• ŷᵢ = predicted label for sample i
• yᵢ = true label for sample i

Range

Accuracy is bounded between 0 and 1, where:

• 0 = worst possible accuracy (all predictions wrong)
• 1 = perfect accuracy (all predictions correct)
• 0.5 = random guessing for balanced binary classification

Arguments

• y_true - Ground truth (correct) labels
• y_pred - Predicted labels, as returned by a classifier

Returns

• The fraction of correctly classified samples (float)

Examples

use scirs2_core::ndarray::array;
use scirs2_metrics::classification::accuracy_score;

let y_true = array![0, 1, 2, 3];
let y_pred = array![0, 2, 1, 3];

let acc = accuracy_score(&y_true, &y_pred).unwrap();
assert!((acc - 0.5).abs() < 1e-10); // 2 out of 4 are correct