Function vectorize 

pub fn vectorize<T, U, F, D>(
    f: F,
) -> impl Fn(&Array<T, D>) -> FerrayResult<Array<U, D>>
where
    T: Element + Copy,
    U: Element,
    D: Dimension,
    F: Fn(T) -> U,

Wrap a scalar function to operate elementwise on arrays of any dimension.

Returns a closure that accepts &Array<T, D> and returns FerrayResult<Array<U, D>>, applying f to every element.

This is NumPy’s np.vectorize — in Rust it is essentially .mapv() wrapped as a reusable callable.

Works with any dimension type: Ix1, Ix2, IxDyn, etc.

Example

let square = vectorize(|x: f64| x * x);
let result = square(&input_array)?;