Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Basically I'm struggling to prove that if you take the indicator function $f$ on $\mathbb{Q}$ intersect $[0,1]$ and $\phi, \psi$ are step functions such that $\phi \leq f \leq \psi$ then $\phi \leq 0$ and $\psi \geq 0$ for all but finitely many $x$

As step functions can only be defined on intervals (by my definition I am working with) I'm trying to describe the places where the indicator function is 1

EDIT: so as $\mathbb{Q}$ is a set of discrete points, when we define a step function on $[0,1]$ can we can not find an interval such that the indicator function of $\mathbb{Q}$ intersect $[0,1]$ where it is 1, so can not find anywhere where $\phi > 0$? Unless we can define step functions pointwise?

share|cite|improve this question
$\mathbb{Q}$ does not contain any interval if that is what you're asking. – HbCwiRoJDp Jan 19 '12 at 16:04
@Karatug Ozan: thank you for confirming that – Freeman Jan 19 '12 at 16:09
+1 Nice question. – user38268 Jan 19 '12 at 16:17
up vote 2 down vote accepted

The result follows immediately from the fact that any step function can be written as

$$g=\sum_{i=1}^n a_i \chi_{I_i} \,,$$

where the intervals $I_i$ are pairwise disjoint. Note that the intervals could actually be trivial (i.e. a single point).

Now if $\phi(x) >0$, then $x \in I_i$ for some $I_i$. But then, if $I_i$ is not trivial, it contains an irrational number $y$, and thus $\phi(y)=\phi(x)>0=f(y)$, contradiction.

The other implication follows from the density of the rationals.

share|cite|improve this answer
Thank you :) this is very helpful – Freeman Jan 21 '12 at 16:52

All you're trying to prove is that, for example, $\phi\geqslant0$ a.e. on $[0,1]$. But, you know that the set of values where this isn't true is contained in $\mathbb{Q}\cap[0,1]$, which then must be countable, which tells us you is measure zero.

P.S. There seems to be some misconception on measure zero, as is suggested by the wording of your question. Being measure zero is not equivalent to containing no intervals, as is easily seen by the fact that $(\mathbb{R}-\mathbb{Q})\cap[0,1]$ has empty interior yet is not measure zero--if it did have measure zero then so would $((\mathbb{R}-\mathbb{Q})\cap[0,1])\cup(\mathbb{Q}\cap[0,1])=[0,1]$!

share|cite|improve this answer
Hmm, if you elaborate that would be very helpful. How do we translate the property of countable to finite? As we are looking for a finite number such that the step function is strictly greater than 0. Does this mean the step function maps some points to points greater than 0? As surely any interval mapping to greater that 0 contains infinite elements? – Freeman Jan 19 '12 at 16:08
I have not yet defined measure in my course, that'll be next lecture sorry :) – Freeman Jan 19 '12 at 16:12

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.