To be fractions that can be represented exactly by floating point
To be fractions that can be represented exactly by floating point