Particular point topology

From Wikipedia, the free encyclopedia

This is a topology where inclusion of a particular point defines openness. Let X be any set and p\in X. A non empty subset of X is open if and only if it contains p. There are a variety of cases which are individually named:

  • If X = {0,1} with p = 0 we call X the Sierpiński space.
  • If X is finite we call the topology on X the Finite Particular Point topology
  • If X is countable we call the topology on X the Countable Particular Point topology
  • If X is uncountable we call the topology on X the Uncountable Particular Point topology

A generalization of the particular point topology is the closed extension topology. In the case when X\backslash \{p\} has the discrete topology, the closed extension topology is the same as the particular point topology.

This topology is used to provide interesting examples and counterexamples.

Contents

[edit] Properties

Empty interior
Every x\in X with x\ne p is a limit point of X. So the closure of any open set other than \emptyset is X. No closed set contains p so the interior of every closed set is \emptyset.

[edit] Connectedness related

Path and locally connected but not arc connected
f(t) = \begin{cases} x & t\in [0, {1 \over 2})  \\ p & t={1 \over 2} \\ y & t\in ({1 \over 2 }, 1] \end{cases}
f is path for all x,y ∈ X. However since p is open the inverse image under a continuous one to one would be an open single point of [0,1] which is a contradiction.
Dispersion point, example of a set with
p is a dispersion point for X. That is X\{p} is totally disconnected.
Hyperconnected but not ultraconnected
Every open set contains p hence X is hyperconnected. But if a,b\in X with p\ne a \ne b\, and\, p\ne b then {a} and {b} are disjoint closed sets and thus X is not ultraconnected. Note of course that if X is the Sierpinski space then no such a and b exist and X is in fact ultraconnected.

[edit] Compactness related

Closure of compact not compact
The set {p} is compact. However its closure (the closure of a compact set) is the entire space X and if X is infinite this is not compact (since any set {t,p} is open). For similar reasons if X is uncountable then we have an example where the closure of a compact set is not a Lindelöf space.
Pseudocompact but not weakly countably compact
First there are no disjoint non-empty open sets (since all open sets contain 'p'). Hence every continuous function must be constant, and hence bounded proving that X is a pseudocompact space. Any set not containing p does not have a limit point thus if X if infinite it is not weakly countably compact.
Locally compact but not strongly locally compact. Both possibilities regarding global compactness.
If x ∈ X then the set {x,p} is a compact neighborhood of x. However the closure of this neighborhood is all of X and hence X is not strongly locally compact.
In terms of global compactness, X finite if and only if X is compact. The first implication is immediate, the reverse implication follows from noting that \bigcup_{x\in X} \{p,x\} is an open cover with no finite subcover.

[edit] Limit related

Limit point but not an accumulation point
A sequence {ai} converges whenever \exists x\in X such that for all but a finite number of the a_i,\,  a_i \ne p \implies a_i = x. An accumulation point will be an b such that infinitely many of the ai = b and note there may be any countable number of b's. Thus any countably infinite set of distinct points forming a sequence does not have an accumulation point but does have a limit point!
Limit point but not a ω-accumulation point
Let Y be any subset containing p. Then for any q\neq p q is a limit point of Y but not a ω-accumulation point. Because this makes no use of properties of Y it leads to often cited counter examples.

[edit] Separation related

T0
Since there are no disjoing open sets except in the Sierpinski space case X is T0 but satisfies no higher separation axioms.
Separability
{p} is dense and hence X is a separable space. However if X is uncountable then X\{p} is not separable. This is an example of a subspace of a separable space not being separable.
Countablility (first but not second)
If X is uncountable then X is first countable but not second countable.
Comparable ( Homeomorphic topology on the same set that is not comparable)
Let p,q\in X with p\ne q. Let t_p = \{S\subset X \,|\, p\in S\} \, and\,  t_q = \{S\subset X \,|\, q\in S\}. That is tq is the particular point topology on X with q being the distinguished point. Then (X,tp) and (X,tq) are homeomorphic incomparable topologies on the same set.
Density (no nonempty subsets dense in themselves)
Let S be a subset of X. If S contains p then S has no limit points (see limit point section). If S does not contain p then p is not a limit point of S. Hence S is not dense if S is nonempty.
Not first category
Any set containing p is dense in X. Hence X is not a union of nowhere dense subsets.

[edit] Sierpiński space

The space S=(X,\mathcal{T}) composed of the set X = {0,1} endowed with the topology \mathcal{T}=\{\varnothing,\{0\},\{0,1\}\} is called the connected two-point set or Sierpiński space. It is named after Wacław Sierpiński. It has several properties unique to the two point case (and hence it has its own name):

  • S is an inaccessible Kolmogorov space; i.e. S satisfies the T0 axiom, but not the T1 axiom.
  • A topological space is Kolmogorov if and only if it is homeomorphic to a subspace of a power of S.
  • For any topological space X with topology T, let C(X,S) denote the set of all continuous maps from X to S, and for each subset A of X, let I(A) denote the indicator function of A. Then the mapping f : T → C(X,S) defined by f(U) = I(U) is a bijective correspondence.
  • If X is a topological space with topology T, then the weak topology on X generated by C(X,S) coincides with T.
  • S is a sober space.
  • S is generated by the prametric d(0,1) = 1 and d(1,0) = 0. That is, S is prametrizable but not metrizable.
  • See also the article above, particular point topology, which lists many other useful facts about Sierpinski space.

The Sierpiński space has important relations to the theory of computation and semantics [1].

[edit] See also

[edit] References

  1. ^ An online paper, it explains the motivation, why the notion of “topology” can be applied in the investigaton of concepts of the computer science. Alex Simpson: Mathematical Structures for Semantics. Chapter III: Topological Spaces from a Computational Perspective. The “References” section provides many online materials on domain theory.