Depending on the source/usage of your BigDecimal
values it might be faster to check if the scale <= 0 first. If it is, then it's definitely an integer value in the mathematical sense. If it is >0, then it could still be an integer value and the more expensive test would be needed.