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

Let $S$ be the set of continuous function $f$ defined on some segment $[0,a_f], a_f \ge 0$, and such that $f(0)=0$.
For $f$ and $g$ in $S$, let $$ c_{fg}=\max\{z : f(x)=g(x) \text{ for all } x \in [0,z] \}, $$ then, define the metric, $$ d(f,g) = a_f - c_{fg} + a_g - c_{fg}, $$ Show the triangle inequality holds for $d$.

The only proof I see is a case by case analysis of all the possibilities.
I don't see how to enumerate all the possibilities, there seems to be too much.
Also, for the case $$ \begin{cases} f,g,h \text{ are different functions} \\ c_{fg} < c_{fh} < c_{hg} \\ a_{fg}<a_{fh}<a_{hg} \end{cases} $$ I don't see why the triangle inequality holds.

share|cite|improve this question
Perhaps it is easier to proceed by contradiction, that is, suppose that $d(f,h)>d(f,g)+d(g,h)$, and see what happens... – Matemáticos Chibchas Feb 1 '13 at 1:25
up vote 1 down vote accepted

Suppose $a_f+a_h-2c_{fh}>a_f+a_g-2c_{fg}+a_g+a_h-2c_{gh}$, that is $c_{fg}+c_{gh}>c_{fh}+a_g$. By properties of the supremum we have

$$c_{fg}+c_{gh}=\sup\bigl\{w+z: f=g\ \style{font-family:inherit;}{\text{on}}\ [0,w], g=h\ \style{font-family:inherit;}{\text{on}}\ [0,z]\bigr\}\,,$$

and so (again by the definition of supremum) there exist $w,z$ as in the set above such that $w+z-a_g>c_{fh}$. Finally, recall that

$$c_{fh}=\sup\bigl\{t: f=h\ \style{font-family:inherit;}{\text{on}}\ [0,t]\bigr\}\,;$$

but $f=g$ on $[0,w]$ and $g=h$ on $[0,z]$, and so $f=h$ on $\bigl[0,\min\{w,z\}\bigr]$, which forces $w+z-a_g>\min\{w,z\}$, or equivalently $\max\{w,z\}>a_g$, which is impossible, since by (tacit) definition we have $w\leq\min\{a_f,a_g\}, z\leq\min\{a_g,a_h\}$, and in particular $w,z\leq a_g$. This concludes the proof.

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.