Function functional_msle 

pub fn functional_msle(
    y_true: &FdMatrix,
    y_pred: &FdMatrix,
    argvals: &[f64],
) -> Result<f64, FdarError>

Functional Mean Squared Logarithmic Error integrated over argvals.

Computes functional_msle = (1/n) * sum_i ∫ (ln(1+y_true_i(t)) - ln(1+y_pred_i(t)))^2 dt where the integral is approximated by Simpson’s rule over argvals.

MSLE is designed for non-negative targets (e.g. counts, prices). Applying it to values below -1 yields undefined logarithms.

Errors

• FdarError::InvalidDimension if shapes are inconsistent.
• FdarError::InvalidParameter if any value of y_true or y_pred is <= -1 (making ln(1+x) undefined).