3.5. Subspace Topology#

Let (X,d) be a metric space. Let YX. Recall from Definition 3.3 that (Y,d) is a metric subspace with the distance function d restricted to Y×Y.

Remark 3.6

Let Y be a metric subspace of X and SY. The interior, closure and boundary of S w.r.t. X and w.r.t. Y may be different.

If a subspace hasn’t been specified, by default, we shall assume that we are computing the interior, closure and boundary w.r.t. the metric space X.

3.5.1. Open Balls#

Theorem 3.38 (Open balls in the metric subspace)

Let (Y,d) be a metric subspace of (X,d). Let BX(p,r) denote an open ball of radius r at pX. Let BY(p,r) denote an open ball of radius r at pY. Then

BY(p,r)=BX(p,r)Y.

Proof. We recall that

BX(p,r)={yX|d(p,y)<r}.

Similarly,

BY(p,r)={yY|d(p,y)<r}.

We first show that BY(p,r)BX(p,r)Y

  1. Let yBY(p,r).

  2. Then yYX and d(p,y)<r.

  3. Hence yBX(p,r)Y.

We now show that BX(p,r)YBY(p,r).

  1. Let yBX(p,r)Y.

  2. Then yY and yX and d(p,y)<r.

  3. Hence yY and d(p,y)<r.

  4. Hence yBY(p,r).

3.5.2. Open Sets#

Theorem 3.39 (Open sets in the metric subspace)

Let (Y,d) be a metric subspace of (X,d). Let SY. Then S is open in (Y,d) if and only if S=OY where O is an open subset of (X,d).

Proof. A subset S of Y is open in (Y,d) if for every xS, there exists an open ball BY(x,r)SY.

Assume that S=OY where O is open in X.

  1. Let xS.

  2. Then xO and xY.

  3. Since O is open in X, hence there is an open ball BX(x,r)O.

  4. Then BX(x,r)YOY=S.

  5. By Theorem 3.38,

    BX(x,r)Y=BY(x,r)

    is an open ball in the metric subspace (Y,d) of radius r>0 around xY.

  6. Hence S is open in (Y,d).

For the converse, assume that S is open in Y.

  1. For every xS, there is an open ball BY(x,rx)S.

  2. Hence xSBY(x,rx)S.

  3. Also, xBY(x,rx) implies that SxSBY(x,rx).

  4. Thus, S=xSBY(x,rx).

  5. By Theorem 3.38

    BY(x,rx)=BX(x,rx)Y

    for every xS.

  6. Define T=xSBX(x,rx).

  7. Then T is a union of open balls of X.

  8. Hence T is open in X.

  9. Also

    S=xSBY(x,rx)=xS(BX(x,rx)Y)=(xSBX(x,rx))Y=TY.
  10. Hence S=TY where T is an open set of (X,d).

Example 3.13

  • [0,1) is open in the metric space R+.

3.5.3. Subspace Topology#

Recall from Definition 3.7 that a topology on a set S is a collection of sets T such that

  1. Empty set and the whole set are elements of T.

  2. T is closed under arbitrary union.

  3. T is closed under finite intersection.

Theorem 3.40 (The subspace topology)

Let (Y,d) be a metric subspace of (X,d). The open sets of the (Y,d) form a topology.

  1. is open in (X,d).

  2. Hence Y= is open in (Y,d).

  3. X is open in (X,d).

  4. Hence XY=Y is open in (Y,d).

  5. Let {Ai|iI} be a family of open sets of (Y,d).

  6. Then Ai=BiY for every iI with Bi open in X.

  7. Let B=Bi. Then B is open in (X,d).

  8. But Ai=(BiY)=(Bi)Y=BY.

  9. Since B is open in (X,d), hence BY is open in (Y,d).

  10. Let {A1,,An} be a finite collection of open sets of (Y,d).

  11. Then Ai=BiY for every i[1,,n] such that Bi is open in (X,d).

  12. Then Ai=(BiY)=(Bi)Y.

  13. Since Bi is open in (X,d) (a finite intersection), hence Ai is open in (Y,d).

3.5.4. Closed Sets#

Theorem 3.41 (Closed sets in subspace topology)

Let Y be a metric subspace of X. Let SY.

S is closed in Y if and only if S=CY where C is a closed subset of X.

Proof. Let S be closed in Y.

  1. Then, YS is open in Y.

  2. By definition of subspace topology,

    YS=YO

    where O is some open subset of X.

  3. Then,

    S=Y(YS)=Y(YO)=YO=Y(XO).
  4. But C=XO is closed in X.