Skip to content
Tutorials Center
My learning

Topics

Lesson 25 of 36 ยท Connectedness

Path-Connectedness

Learn path-connectedness, why it implies connectedness, why the converse fails (topologist's sine curve), and test regions of a grid by code.

Connectedness says a space has no separation. That is a clean definition, but it is not the first thing most people imagine when they think "one piece". The everyday picture is: you can walk from any point to any other without leaving the set. Topology names this idea path-connectedness. It is a stronger property than connectedness, usually easier to check, and in most nice spaces the two agree.

You will define it, prove the implication, see the famous example where it fails, and write a small program that checks it on a pixel grid.

Paths

A path in a space X from p to q is a continuous function ฮณ: [0, 1] โ†’ X with ฮณ(0) = p and ฮณ(1) = q. Think of t as time and ฮณ(t) as the position. The path does not have to be smooth or injective; it just has to be continuous.

A space X is path-connected if every two points can be joined by a path in X.

Easy examples

  • Every interval of โ„ is path-connected: use ฮณ(t) = (1 โˆ’ t)p + tq.
  • Every convex subset of โ„โฟ is path-connected, by the same straight-line formula.
  • โ„ยฒ \ {0} is path-connected: go around the hole.
  • โ„ \ {0} is not, since a path from โˆ’1 to 1 would have to cross 0 (the IVT).
  • A circle is path-connected: walk along the arc.

Path-connected implies connected

Theorem. Every path-connected space is connected.

Proof. Suppose X = U โˆช V is a separation. Pick p โˆˆ U and q โˆˆ V and a path ฮณ from p to q. Then ฮณโปยน(U) and ฮณโปยน(V) are disjoint open sets covering [0, 1], containing 0 and 1 respectively, so they are nonempty. That is a separation of [0, 1], which is connected. Contradiction. โˆŽ

The proof imports connectedness of the interval, which is why the interval is the standard "test path".

Python Path-connectedness of regions on a grid (a discrete stand-in)

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The program joins two cells only when you can step from one to the other, which is the discrete version of "there is a path in the set". The region has three pieces. A real set would need real continuous paths, but the idea is the same.

Connected does not imply path-connected

The standard counterexample is the topologist's sine curve:

S = { (x, sin(1/x)) : 0 < x โ‰ค 1 } โˆช { (0, y) : โˆ’1 โ‰ค y โ‰ค 1 }.

  • It is connected. The graph part G = {(x, sin(1/x)) : 0 < x โ‰ค 1} is the continuous image of the interval (0, 1], so it is connected. The vertical segment L = {0} ร— [โˆ’1, 1] consists of limit points of G: every point (0, y) is a limit of points on the graph, because sin(1/x) takes the value y for arbitrarily small x. So G โІ S โІ cl(G), and a set between a connected set and its closure is connected.
  • It is not path-connected. There is no path from a point of G to a point of L. Roughly: a continuous path approaching the segment has to follow the graph, whose height oscillates infinitely often between โˆ’1 and 1 as x โ†’ 0. By the IVT its y-coordinate would have to oscillate infinitely often in a bounded time, contradicting continuity at the endpoint.
Python How fast sin(1/x) oscillates as x approaches 0

Press Run (or Ctrl + Enter) to see the result here.

Program input ยท what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

The number of sign changes grows without bound as the interval starts closer to 0. A path arriving at x = 0 would have to complete infinitely many full oscillations in finite time, which a continuous function of a single variable approaching a point cannot do. That is the intuition behind "no path".

When the two notions agree

If a space is connected and locally path-connected (every point has arbitrarily small path-connected neighborhoods), then it is path-connected. This covers every open connected subset of โ„โฟ, every manifold and every CW complex, so for nearly everything you meet in practice, "connected" and "path-connected" are interchangeable. The topologist's sine curve is deliberately pathological.

When you need to prove a space is connected, look for paths first. When you need to prove it is not connected, look for a separation, or show a continuous function onto {0, 1}.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. What is a path from p to q in a space X?

    A path only needs to be continuous; it need not be straight, smooth or injective.
  2. 2. Which statement is true for every topological space?

    A separation of X would pull back along a path to a separation of [0,1], which is connected. The converse fails for the topologist's sine curve.
  3. 3. Why is the topologist's sine curve connected?

    A set lying between a connected set and its closure is connected. The vertical segment consists of limit points of the graph.
  4. 4. Which set is NOT path-connected?

    A path from โˆ’1 to 1 in โ„ would cross 0 by the intermediate value theorem. The other sets allow paths between any two points.

Summary

  • A path is a continuous map from [0, 1]; path-connected means any two points are joined by a path.
  • Path-connected implies connected; the topologist's sine curve shows the converse fails.
  • Convex sets and open connected subsets of โ„โฟ are path-connected.
  • For locally path-connected spaces, connected and path-connected coincide.
  • Grid searches (breadth-first) are the discrete analogue of looking for paths.

Next: Components and Local Connectedness