This function rounds its argument. The first argument is the number to round, and the optional second argument specifies the number of fractional digits to round to.
The function rounds to integer value if the second argument is missing.
The return value is long if the number of decimal places to keep is zero, otherwise the return value is double.
Negative value for the number of decimal places results rounding to integer value.
ROUND(undef) is undef or raises an error if the option RaiseMatherror is set in bit sbMathErrUndef.
Also ROUND(val,undef) is equivalent to ROUND(value).