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

It is a well known fact, that if $f:X \to \mathbb{R}$ (where $X \subset \mathbb{R}$) is monotonic, and $a \in X^+, b\in X^-$, where $$X^+ = \left\{ {x\in X:\forall \varepsilon > 0,\;\;\left( {x - \varepsilon ,x} \right) \cap X \ne \varnothing} \right\}$$ $$X^- = \left\{ {x\in X:\forall \varepsilon > 0,\;\;\left( {x,x + \varepsilon } \right) \cap X \ne \varnothing} \right\},$$ then $$\lim_{x \to a^+} f\left( x \right),\qquad\lim_{x \to b^-} f\left( x \right)$$ exist. My question is, given such a function, and a point $a \in X^+$, does there exist an $\varepsilon > 0$ such that $$ f\left| {_{\left( {x - \varepsilon ,x} \right) \cap X} } \right. $$ is continuous (doesn't have jumps). I want to know if this fact is true, to prove that the set of discontinuities of a monotonic function is countable (because in this case, to each such point, I associate to it, an open set, and therefore a rational number).

The only possible counterexample is a function that has a dense set of jumps. Please don't give me proof about the other proof, I only know this, because I know other proof without using this fact, but I think that this is also true.

share|cite|improve this question
Apologies if I changed your intended meaning in any way; please feel free to make further edits if mine aren't to your liking. – Zev Chonoles Oct 9 '11 at 3:35
If I understand you correctly, then $f(x)=x+\frac{1}{\lfloor 1/x\rfloor}$ on $x\in(-\infty,0)$ is a counterexample for $a=0$. (And its jump points are far from dense). – Henning Makholm Oct 9 '11 at 3:40
Are you sure you have your limits correct? It seems to me that if $a$ lies in the set you describe as $X^+$, then you can define the limit as $x$ approaches $a$ from the left of $f(x)$, but in general $\lim\limits_{x\to a^+}f(x)$ need not exist (just take $X=[0,1]$; then $1$ is in $X^+$, but the limit from the right does not exist, because the function is not defined to the right of $1$). – Arturo Magidin Oct 9 '11 at 3:47
up vote 1 down vote accepted

Proposition 1. A monotone function $f: \mathbb{R} \to \mathbb{R}$ can only have jump discontinuities.

Corollary 2. A monotone function $f: \mathbb{R} \to \mathbb{R}$ has at most countably many discontinuities.

Proposition 3. For any countable $A \subset \mathbb{R}$ there exists a bounded increasing function $f: \mathbb{R} \to \mathbb{R}$ such that $f$ is continuous precisely on $\mathbb{R}\setminus A$.

Only the last statement is moderately difficult to prove.

share|cite|improve this answer

I think you have your definitions of $X^+$ and $X^-$ the wrong way around: under your definition, if $X=[0,1]$, then $1\in X^+$, but $\lim\limits_{x\to 1^+} f(x)$ does not make sense, so it cannot exist.

Your question has a negative answer.

Let $\{q_n\}$ be an enumeration of the rationals on $[0,1]$; that is, $q_n$ is a rational on $[0,1]$ for each $n$, and for every rational $q$ in $[0,1]$ there exists one and only one $n$ such that $q=q_n$.

Define $f\colon [0,1]\to \mathbb{R}$ by $$f(x)= \sum_{q_n\gt x}\frac{1}{2^n}.$$ Then $f(x)$ is monotonic: if $x\leq y$, then $\{n\in\mathbb{N}\mid q_n\gt y\}\subseteq \{n\in\mathbb{N}\mid q_n\leq x\}$, so $f(x)\geq f(y)$.

Moreover, $f$ is discontinuous from the left at every rational number: if $q\in (0,1]\cap\mathbb{Q}$, there exists $n$ such that $q=q_n$. Let $\epsilon=\frac{1}{2^{n+1}}$. Then for every $x\in(0,q)$, $f(x)-f(q) \geq \frac{1}{2^n}\gt \epsilon$, so $f$ is not continuous from the left at $q$.

In particular, for every $x$, $1\geq x\gt 0$ and every $\epsilon\gt 0$, $f$ is not continuous on $(\max(x-\epsilon,0),x)$.

The function is continuous from the right at every rational, but it is likewise true that for all $x$, $ 0\leq x\lt 1$, and for every $\epsilon\gt 0$, $f$ is not continuous on $(x,\min(x+\epsilon,1))$.

Note that in this case, $X^+$ and $X^-$ are $[0,1)$ and $(0,1]$ (in some order, depending on the correct definitions of $X^+$ and $X^-$).

share|cite|improve this answer

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.