logo

거리공간에서 연속함수일 동치 조건 📂MetricSpace

거리공간에서 연속함수일 동치 조건

Theorem 1

For two metric spaces (X,dX)(X,d_{X}), (Y,dY)(Y,d_{Y}), let EXE\subset X and pEp \in E, f:EYf : E \to Y. Then, the following three statements are equivalent.

(1a) ff is continuous at pp.

(1b) limxpf(x)=f(p) \lim \limits_{x \to p} f(x)=f(p).

(1c) For limnpn=p\lim \limits_{n\to\infty} p_{n}=p, {pn}\left\{ p_{n} \right\}, it follows that limnf(pn)=f(p)\lim \limits_{n\to\infty} f(p_{n})=f(p).

Proof

Theorem 2

For two metric spaces (X,dX)(X,d_{X}), (Y,dY)(Y,d_{Y}), let f:XYf : X \to Y. Then the following three statements are equivalent.

(2a) ff is continuous on XX.

(2b) For every open set YY OYO_{Y}, f1(OY)f^{-1}(O_{Y}) is an open set in XX.

(2c) For every closed set YY CYC_{Y}, f1(CX)f^{-1}(C_{X}) is a closed set in XX.

Here, f1f^{-1} denotes the preimage rather than the inverse function.

Proof

  • (2a)     \implies (2b)

    Assume ff is continuous on XX. Let OYO_{Y} be an open set in YY. By the definition of an open set, it suffices to show that every point in f1(OY)f^{-1}(O_{Y}) is an interior point of f1(OY)f^{-1}(O_{Y}). Consider an arbitrary f(p)OYf(p) \in O_{Y}. Then pf1(OY)p \in f^{-1}(O_{Y}) holds. Since OYO_{Y} is open, f(p)f(p) is an interior point of OYO_{Y}. Thus,

    dY(y,f(p))<ε    yOY \begin{equation} d_{Y}(y,f(p)) < \varepsilon \implies y \in O_{Y} \label{eq1} \end{equation}

    holds for some positive ε\varepsilon. Then, since ff is continuous on XX, for this ε\varepsilon, there exists

    dX(x,p)<δ    dY(f(x),f(p))<ε \begin{equation} d_{X}(x,p) < \delta \implies d_{Y}(f(x),f(p))<\varepsilon \label{eq2} \end{equation}

    some δ>0\delta >0 such that it holds. However, by (eq1)\eqref{eq1}, (eq2)\eqref{eq2},

    dX(x,p)<δ    dY(f(x),f(p))<ε    f(x)OY d_{X}(x,p) < \delta \implies d_{Y}(f(x),f(p))<\varepsilon \implies f(x)\in O_{Y}

    holds, and hence xf1(OY)x \in f^{-1}(O_{Y}). Therefore, for some positive δ\delta,

    dX(x,p)<δ    xf1(OY) d_{X}(x,p) < \delta \implies x \in f^{-1}(O_{Y})

    holds, so pp is an interior point of f1(OY)f^{-1}(O_{Y}) and f1(OY)f^{-1}(O_{Y}) is open.

  • (2b)     \implies (2a)

    Assume (2b). Choose arbitrary pXp \in X and ε>0\varepsilon >0. Let the set OYO_{Y} be defined as follows.

    OY={y:dY(y,f(p))<ε} O_{Y} =\left\{ y : d_{Y}(y,f(p))<\varepsilon \right\}

    Then, OYO_{Y} is an open set in YY. Hence, by assumption, f1(OY)f^{-1}(O_{Y}) is open in XX. Therefore,

    dX(x,p)<δ    xf1(OY) d_{X}(x,p) <\delta \implies x \in f^{-1}(O_{Y})

    there exists a positive δ>0\delta >0 such that this holds. Then,

    xf1(OY)anddX(x,p)<δ    dY(f(x),f(p))<ε x\in f^{-1}(O_{Y}) \quad \text{and} \quad d_{X}(x,p)< \delta \implies d_{Y}(f(x),f(p))<\varepsilon

    holds, so by the definition of continuity, ff is continuous at every pXp \in X.

  • (2b)     \iff (2c)

    By showing (2b)     \implies (2c), the opposite direction can be demonstrated with similar logic, hence the remaining proof is omitted. Assume (2b) holds. Unpacking for the open set YY OYO_{Y} yields the following.

    OY is open in Y    f1(OY) is open in X O_{Y}\mathrm{\ is\ open\ in\ } Y \implies f^{-1}(O_{Y})\mathrm{\ is\ open\ in\ }X

    As the complement of an open set is a closed set, let OY=(CY)cO_{Y}=(C_{Y})^{c}, then the statement becomes:

    CY is closed in Y    f1((CY)c) is open in X C_{Y}\mathrm{\ is\ closed\ in\ } Y \implies f^{-1}((C_{Y})^{c})\mathrm{\ is\ open\ in\ }X

    However, since f1((CY)c)=(f1(CY))cf^{-1}((C_{Y})^{c})=(f^{-1}(C_{Y}))^{c}, the following holds:

    CY is closed in Y    (f1(CY))c is open in X C_{Y}\mathrm{\ is\ closed\ in\ } Y \implies (f^{-1}(C_{Y}))^{c}\mathrm{\ is\ open\ in\ }X

    Additionally, since the complement of an open set is closed, the following holds:

    CY is closed in Y    f1(CY) is closed in X C_{Y}\mathrm{\ is\ closed\ in\ } Y \implies f^{-1}(C_{Y})\mathrm{\ is\ closed\ in\ }X