Here's an explicit mapping, although another kind of proof might be more elegant.
By "e", I mean the famous irrational number e, but any irrational number will do.
For any real number x, compute f(x) in the following way:
If x can be written in the form x=ke+r, where k is a nonnegative integer and r is rational, then set f(x)=x+e. Otherwise, f(x)=x.
First, we show that this function sends R into R\Q:
Suppose that f(x) is rational for some x. If x can be written as x=ke+r, then f(x)=ke+r+e and is rational, from which we can show that e is rational (note that k is nonnegative). Contradiction. On the other hand, if x is not of the form ke+r, then f(x)=x; this implies that x is rational, so x=0e+r. Contradiction again.
Thus, f(x) is not rational.
Next, let's show that f is one-to-one.
Suppose that f(x)=f(y) for some x,y in R. If f(x) and f(y) are evaluated by the same rule, it's easy to show that x=y. So, let's suppose x=ek+r, while y cannot be written in that form. Then f(x)=ek+r+e, while f(y)=y; we now have y=f(y)=f(x)=(k+1)e+r, which contradicts our assumption that y cannot be written in that form.
So far, we've shown that f is an injection from R into R\Q. Shall we go for bijection? Why not.
Let z be a member of R\Q. If z-ne is rational for some nonnegative integer n, then n>0 (for otherwise z is in Q). We have z=ne+r (where r is rational), and z=f(z-e). If, on the other hand, z-ne is *not* rational for any nonnegative integer n, then f(z)=z.
So, the given function is a bijection between R and R\Q.
2007-01-15 07:18:10
·
answer #1
·
answered by Doc B 6
·
1⤊
0⤋