33b7ff128f
Added which number is computed in compute_float. The original comment was very elaborate but ultimately did not mention at all what is being computed using parameters `w, q`, only referencing an external article for the algorithm.