{{alias}}( x, value, dims, k )
    Fills a specified diagonal of a matrix (or stack of matrices) with a
    scalar value.

    The order of the dimension indices contained in `dims` matters. The first
    element specifies the row-like dimension. The second element specifies the
    column-like dimension.

    Each provided dimension index must reside on the interval [-ndims, ndims-1].

    The diagonal offset `k` is interpreted as `column - row`. Accordingly, when
    `k = 0`, the function fills the main diagonal; when `k > 0`, the function
    fills a diagonal above the main diagonal; and when `k < 0`, the function
    fills a diagonal below the main diagonal.

    The function mutates the input ndarray in-place.

    Parameters
    ----------
    x: ndarray
        Input array.

    value: any
        Scalar value. Must be able to safely cast to the input ndarray data
        type. Scalar values having floating-point data types (both real and
        complex) are allowed to downcast to a lower precision data type of the
        same kind (e.g., a scalar double-precision floating-point number can be
        used to fill a 'float32' input ndarray).

    dims: ArrayLikeObject<integer>
        Dimension indices defining the plane in which to fill the diagonal.

    k: integer
        Diagonal offset.

    Returns
    -------
    out: ndarray
        Input array.

    Examples
    --------
    > var x = {{alias:@stdlib/ndarray/zeros}}( [ 3, 3 ] )
    <ndarray>
    > var out = {{alias}}( x, 1.0, [ 0, 1 ], 0 )
    <ndarray>[ [ 1.0, 0.0, 0.0 ], [ 0.0, 1.0, 0.0 ], [ 0.0, 0.0, 1.0 ] ]
    > var bool = ( out === x )
    true

    See Also
    --------