Quant Academy Math · Probability · Quant Finance
Phase 6 - Lesson 6.1

Metric Spaces, Sequences, Completeness, and Compactness

The topology of convergence: distance, Cauchy sequences, why the reals are complete, and the two faces of compactness.

⏱ 55 min● Advanced🔗 Prereqs: 1.x proof, 2.x limits
↖ Phase 6 hub
Builds on: Phase 2 used limits informally; here convergence gets its topological definition.
Leads to: Completeness underlies \(L^p\) spaces (6.5) and the existence of stochastic limits (Phase 8).

Learning Objectives

Click a status chip to cycle: Not started → In progress → Studied → Practiced → Needs review → Mastered.

Key Vocabulary

Metric
A function \(d(x,y)\) that is nonnegative, symmetric, zero iff \(x=y\), and obeys the triangle inequality.
Open set
A set in which every point has a ball entirely inside it; the basis of topology.
Cauchy sequence
A sequence whose terms get arbitrarily close to each other: \(d(x_m,x_n)\to0\).
Complete space
A metric space in which every Cauchy sequence converges to a point of the space.
Compact
A set for which every open cover has a finite subcover; in \(\mathbb{R}^n\), closed and bounded.
Continuity
A map with \(x_n\to x\Rightarrow f(x_n)\to f(x)\); equivalently preimages of open sets are open.

Intuition & Motivation

Intuition
Analysis is the study of limits, and a limit needs two things: a notion of closeness and a guarantee that sequences trying to converge actually land somewhere. A metric supplies closeness. Completeness supplies the landing: the rationals have holes (a sequence homing in on \(\sqrt2\) is Cauchy but converges to nothing rational), while the reals are built precisely to fill them. Compactness is the property that packs an infinite set tightly enough that sequences cannot escape - it is why a continuous function on a closed bounded set attains its maximum.

Metrics and open sets

Definition - Metric space
A set \(X\) with \(d:X\times X\to[0,\infty)\) such that (i) \(d(x,y)=0\iff x=y\), (ii) \(d(x,y)=d(y,x)\), and (iii) \(d(x,z)\le d(x,y)+d(y,z)\) (triangle inequality).

The open ball \(B(x,r)=\{y:d(x,y)\lt r\}\) generates the topology: a set is open if it contains a ball around each of its points, and closed if its complement is open.

Convergence, Cauchy, completeness

A sequence converges, \(x_n\to x\), if \(d(x_n,x)\to0\). It is Cauchy if for every \(\varepsilon\gt 0\) there is \(N\) with \(d(x_m,x_n)\lt \varepsilon\) for all \(m,n\ge N\). Every convergent sequence is Cauchy; the converse defines completeness.

Theorem - Cauchy but not convergent in \(\mathbb{Q}\)
The sequence \(x_{n+1}=\tfrac12(x_n+2/x_n)\) of rationals is Cauchy but its limit \(\sqrt2\) is irrational, so \(\mathbb{Q}\) is not complete. Completing \(\mathbb{Q}\) (filling all such gaps) yields \(\mathbb{R}\).
Proof
Newton’s iteration for \(x^2-2\) contracts: the errors satisfy \(|x_{n+1}-\sqrt2|\le C|x_n-\sqrt2|^2\), so terms bunch together (Cauchy). Yet no rational squares to exactly 2, so within \(\mathbb{Q}\) there is no limit. ∎

Compactness

Theorem - Heine–Borel & sequential compactness
In \(\mathbb{R}^n\), a set is compact (every open cover has a finite subcover) iff it is closed and bounded iff it is sequentially compact (every sequence has a convergent subsequence with limit in the set).

Compactness is what makes optimization well-posed: a continuous \(f\) on a compact \(K\) attains its supremum and infimum - the extreme value theorem - because a maximizing sequence has a convergent subsequence landing in \(K\).

Worked Example - The closed bounded set matters
1
On \((0,1]\) (bounded, not closed) the continuous \(f(x)=1/x\) is unbounded - no maximum. Compactness fails because the point 0 is missing.
2
On \([0,\infty)\) (closed, not bounded) the continuous \(f(x)=x\) has no maximum. Compactness fails for lack of boundedness.
3
On \([0,1]\) (closed and bounded, hence compact) every continuous function attains a max and a min.
4
Moral: both halves of ‘closed and bounded’ are load-bearing; drop either and extrema can vanish.

Interactive: a Cauchy sequence filling a hole in \(\mathbb{Q}\)

Iterate the rational recursion toward \(\sqrt2\) and watch consecutive terms bunch together - Cauchy in \(\mathbb{Q}\), but the limit escapes \(\mathbb{Q}\).

Common Mistakes to Avoid
  • Assuming every Cauchy sequence converges - only true in a complete space.
  • Thinking ‘bounded’ alone gives compactness; you also need closedness (in \(\mathbb{R}^n\)).
  • Confusing an open cover having a finite subcover (compactness) with the set merely being finite.
  • Using continuity to conclude a maximum exists on a non-compact domain.
Quant Practitioner Tips
  • To prove a limit exists without knowing its value, show the sequence is Cauchy in a complete space.
  • Heine–Borel is a \(\mathbb{R}^n\) luxury; in infinite dimensions closed+bounded need NOT be compact.
  • Compactness turns ‘for all \(\varepsilon\)’ arguments into finite ones - the finite subcover is the workhorse.
  • Continuity + compact domain \(\Rightarrow\) uniform continuity and attained extrema; lean on this in analysis proofs.

Knowledge Check

Q1 Medium
Why is \(\mathbb{Q}\) not complete?
It is uncountable
It has a Cauchy sequence (e.g. approaching \(\sqrt2\)) whose limit is not rational
It is not bounded
It has no open sets
Q2 Easy
In \(\mathbb{R}^n\), a set is compact if and only if it is:
Open and bounded
Closed and bounded
Closed and connected
Bounded and countable
Q3 Medium
A continuous function on a compact set is guaranteed to:
Be differentiable
Attain its maximum and minimum
Be linear
Have no zeros

Practical Exercise

Prove that every convergent sequence in a metric space is Cauchy, and explain with one example why the converse can fail.

▶ Show full solution

Suppose \(x_n\to x\). Given \(\varepsilon\gt 0\), pick \(N\) with \(d(x_n,x)\lt \varepsilon/2\) for \(n\ge N\). Then for \(m,n\ge N\), the triangle inequality gives \(d(x_m,x_n)\le d(x_m,x)+d(x,x_n)\lt \varepsilon\), so the sequence is Cauchy.

The converse fails in an incomplete space: in \(\mathbb{Q}\) the sequence \(1,1.4,1.41,1.414,\dots\) is Cauchy but has no rational limit. Completeness is exactly the extra hypothesis that repairs the converse.

After the reveal, answer for yourself: Where in the proof did you use the triangle inequality, and could it work without it?

Lesson Summary

A metric gives closeness; open/closed sets give topology; Cauchy sequences try to converge and completeness guarantees they succeed - the property that distinguishes \(\mathbb{R}\) from \(\mathbb{Q}\). Compactness (closed and bounded in \(\mathbb{R}^n\)) tightly confines sequences, making continuous extrema exist. These are the load-bearing tools for all of analysis and probability.

Retrieval Practice

Close the lesson and answer from memory before checking. This is deliberate, effortful recall - the single highest-yield study action.

▶ Show retrieval prompts & answers
Q: State completeness and why \(\mathbb{R}\) has it but \(\mathbb{Q}\) does not.
A: Completeness: every Cauchy sequence converges in the space. \(\mathbb{R}\) is built to fill the gaps that leave rational Cauchy sequences (e.g. toward \(\sqrt2\)) without a limit.
Q: What does Heine–Borel say and why does it matter?
A: In \(\mathbb{R}^n\), compact \(\iff\) closed and bounded; compactness guarantees continuous functions attain extrema and turns infinite covers into finite ones.

Completion Checklist

Confidence / mastery rating
Personal notes

Source References

This lesson synthesizes and paraphrases concepts from the sources below. No copyrighted text, problem sets, or solutions are reproduced. Return to the originals for full depth.

← 5.5Next: 6.2 →