{{alias}}( dtype )
    Returns the smallest positive normal value capable of being represented by a
    numeric real type.

    The following numeric real types are supported:

    - float64: double-precision floating-point numbers
    - float32: single-precision floating-point numbers
    - float16: half-precision floating-point numbers

    Parameters
    ----------
    dtype: string
        Numeric type.

    Returns
    -------
    out: number
        Smallest finite normal value.

    Examples
    --------
    > var m = {{alias}}( 'float16' )
    0.00006103515625
    > m = {{alias}}( 'float32' )
    1.1754943508222875e-38

    See Also
    --------