Rounding occurs when you convert a number of greater precision into a number of lesser precision. For instance, when converting a floating-point number (single precision, double precision, or variant data types) into an integer or long data type number. The possible ways of numeric rounding are discussed below.