Help for this page
use Math::BigRat; # Useless use bignum; # Uses M::BigInt and M::BitFloat
use bignum upgrade => "Math::BigRat";
use bigrat;