pub fn max_error(y_true: &[f32], y_pred: &[f32]) -> f32
Maximum residual error max|yᵢ − ŷᵢ|, matching sklearn.metrics.max_error.
max|yᵢ − ŷᵢ|
sklearn.metrics.max_error
Panics if y_true and y_pred differ in length.
y_true
y_pred