Compact iff complete and totally bounded

In metric spaces, compactness is equivalent to completeness plus total boundedness
Compact iff complete and totally bounded

Let (X,d)(X,d) be a and let KXK\subseteq X with the subspace metric.

Theorem: The following are equivalent:

This theorem is the fundamental metric characterization of compactness: “no missing limits” (completeness) plus “finite ε\varepsilon-approximation at every scale” (total boundedness).

Proof sketch: (\Rightarrow) If KK is compact, then KK is complete (every has a , and hence the full sequence converges) and totally bounded (the balls {HAHAHUGOSHORTCODE558s7HBHB(x,ε):xK}\{ (x,\varepsilon):x\in K\} form an ; take a finite subcover).

(\Leftarrow) If KK is totally bounded, then . If KK is complete, that Cauchy subsequence converges to a point in KK. Hence KK is . In metric spaces, sequential compactness is equivalent to compactness.