Calculate the sum of absolute values (L1 norm) of double-precision floating-point strided array elements, ignoring NaN values and using ordinary recursive summation. - View it on GitHub