Compute the range of a one-dimensional double-precision floating-point ndarray, ignoring `NaN` values. - View it on GitHub