# Riemann Integrable

1. Feb 6, 2006

Is it possible to show by induction that f:[a,b]->R, a bounded function, is Riemann integrable if f has a countable number of discontinuities? I'm told this is usually done with Lebesgue integrals, but I don't see why an inductive proof of this using Riemann/Darboux integrals can't work.

2. Feb 6, 2006

### matt grime

What are you inducting on? It can't be the unmber of discontinuities. showing something for a finite number of things in no way says anything about the infinite case.

3. Feb 6, 2006

### NateTG

No.

Consider, for example, the characteristic function of the irrationals.
f(x) is 1 if x is irrational, and 0 if x is rational.

This has a countable number of discontinuities -- one at every rational number -- but is most certainly not Riemann integrable as the lower sum will always be 0, and the upper sum will always be 1.

Last edited: Feb 6, 2006
4. Feb 6, 2006

### matt grime

A 'rational' number of discontinuities? No, it has an uncountable number of discontinuities. It is discontinuous at every point at of the interval [0,1]

5. Feb 6, 2006

Well, f is integrable if it has 1 discontinuity, and whenever it has n discontinuities and is integrable, it implies that having n+1 discontinuities is integrable.

6. Feb 6, 2006

### matt grime

and that tells you nothing (as given) about countably (but not finitely) many discontinuities. (Proof... well, according to this logic, the set of natural numbers is finite by 'induction')

7. Feb 6, 2006