Module scoring 

Functional scoring metrics — MAE, MSE, MAPE, MSLE, explained variance.

All metrics integrate the pointwise error function over argvals using Simpson’s rule, producing a single scalar score per metric. Each metric averages over all curves (rows of the input matrices).

Shape Contract (all five functions)

• y_true.shape() == y_pred.shape() — else InvalidDimension { parameter: "y_pred" }
• argvals.len() == y_true.ncols() — else InvalidDimension { parameter: "argvals" }
• y_true.nrows() >= 1 and argvals.len() >= 2 — else InvalidDimension { parameter: "data" }

Functions

functional_explained_variance

Functional Explained Variance Score integrated over argvals.

functional_mae

Functional Mean Absolute Error integrated over argvals.

functional_mape

Functional Mean Absolute Percentage Error integrated over argvals.

functional_mse

Functional Mean Squared Error integrated over argvals.

functional_msle

Functional Mean Squared Logarithmic Error integrated over argvals.