There doesn't seem to be any compelling reason to use float. Using double is simpler, and avoids otherwise unnecessary float<->double conversions.
There doesn't seem to be any compelling reason to use float. Using double is simpler, and avoids otherwise unnecessary float<->double conversions.