It is for a long time now known how to implement arbitrary precision numbers.
Why is there not some way that I can choose types for my variables as natural number or rational number (a division between arbitrary precision natural numbers) or real number?
What I want to have is this mathematical hierarchy of numbers.