You can also do this:
(Current*100)/Max
That way you don't have to worry about FP inaccuracies if all you need is the integer result.
(Current*100)/Max
Print (200000*100)/500000
That way you don't have to worry about FP inaccuracies if all you need is the integer result.