Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

If a metrizable topological space is totally bounded with one metric, is it totally bounded with all others? A related, stronger question: if every metrization of a topological space is bounded, are they all totally bounded?

share|improve this question
2  
To paraphrase GEdgar: Look for a counterexample for the first question. I answered the second question here: a metrizable space is compact if and only if every compatible metric is bounded. –  t.b. Oct 16 '11 at 23:37

3 Answers 3

up vote 4 down vote accepted

The answer to your first question is "no". The real numbers are homeomorphic to the open interval $(-\frac{\pi}{2},\frac{\pi}{2})$, via the bijection given by $x\mapsto \arctan(x)$. Both spaces have a metrizable topology. But $(-\frac{\pi}{2},\frac{\pi}{2})$ is totally bounded, whereas $\mathbb{R}$ is not. Equivalently, endow $\mathbb{R}$ with the metric $d(a,b) = |\arctan(a)-\arctan(b)|$; this metric induces the same topology on $\mathbb{R}$ as the usual metric, but whereas $\mathbb{R}$ is bounded under $d$, it is not bounded under its usual metric.

Second question was answered inter alia elsewhere.

share|improve this answer

Consider the positive integers with the discrete topology. On the one hand this topology is induced by the usual metric, which is certainly not totally bounded. On the other hand, it is also induced by the metric $$d(n,m)=\left|\frac1n-\frac1m\right|\;,$$ which clearly is totally bounded.

This is really just the observation that $\left\{\dfrac1n:n\in\mathbb{Z}^+\right\}$ and $\mathbb{Z}^+$ are homeomorphic as subspaces of $\mathbb{R}$ with the Euclidean topology, via the map $n\mapsto\dfrac1n$, where the first subspace has compact closure in $\mathbb{R}$.

share|improve this answer

Others have pointed out that total boundedness is not a topological property; what's also worth noting is that it is a uniform property. A uniform space is a space that carries some additional structure beyond its topology that allows you to make sense of such notions as uniform continuity, uniform convergence, total boundedness, Cauchy sequences/nets/filters, completeness, completion...

I'll leave the precise definition to the linked article (there are several equivalent formulations), but the point is that a metric space (more generally, pseudometric space) naturally carries the structure of a uniform space (it's kind of the obvious motivating example -- the other motivating example being topological groups), and that total boundedness is a uniform property. (Worth noting is that so is completeness, and the theorem that a metric space is compact iff it's complete and totally bounded works also for more general uniform spaces. So do a lot of the basic metric space theorems, really, though unfortunately not all of them.)

(By contrast, boundedness isn't even a uniform property, since every metric space, bounded or not, is uniformly homeomorphic to a bounded metric space.)

share|improve this answer

Your Answer

 
discard

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.