Properties of Continuous Functions in Metric Spaces
📂MetricSpace Properties of Continuous Functions in Metric Spaces Theorem: Real Functions Let two functions f f f , g g g be functions from a metric space X X X to the complex numbers.
f : X → C , g : X → C
f:X \to \mathbb{C},\quad g:X \to \mathbb{C}
f : X → C , g : X → C
If the two functions are continuous, then f + g f+g f + g , f g fg f g , f / g f/g f / g are also continuous. However, in the last case, it only holds for g ( x ) ≠ 0 g(x)\ne 0 g ( x ) = 0 being x ∈ X x\in X x ∈ X .
Proof Lemma 1
Let ( X , d ) (X,d) ( X , d ) be a metric space, E ⊂ X E\subset X E ⊂ X a subset, and p p p an accumulation point of E E E . Suppose two complex-valued functions defined on E E E , f : E → C f:E\to \mathbb{C} f : E → C , g : E → C g: E\to \mathbb{C} g : E → C , are given. And suppose the two functions have the following limits at p p p .
lim x → p f ( x ) = A and lim x → p g ( x ) = B
\lim \limits_{x \to p}f(x)=A \quad \text{and} \quad \lim \limits_{x \to p}g(x)=B
x → p lim f ( x ) = A and x → p lim g ( x ) = B
Then
lim x → p ( f + g ) ( x ) = A + B \lim \limits_{x \to p}(f+g)(x)=A+B x → p lim ( f + g ) ( x ) = A + B
lim x → p ( f g ) ( x ) = A B \lim \limits_{x \to p}(fg)(x)=AB x → p lim ( f g ) ( x ) = A B
lim x → p ( f g ) ( x ) = A B , B ≠ 0 \lim \limits_{x \to p}\left( \frac{f}{g} \right)(x)=\frac{A}{B},\ B\ne 0 x → p lim ( g f ) ( x ) = B A , B = 0
Lemma 2
For two ( X , d X ) (X,d_{X}) ( X , d X ) , ( Y , d Y ) (Y,d_{Y}) ( Y , d Y ) , let E ⊂ X E\subset X E ⊂ X and assume p ∈ E p \in E p ∈ E , f : E → Y f : E \to Y f : E → Y . Then the following two conditions are equivalent.
It holds by Lemma 1 and Lemma 2.
■
Theorem: Vector Functions Let f 1 f_{1} f 1 , f 2 f_{2} f 2 , ⋯ \cdots ⋯ , f k f_{k} f k be functions from a metric space X X X to the real numbers. And let f \mathbf{f} f be defined as follows.
f : X → R k and f ( x ) = ( f 1 ( x ) , ⋯ , f k ( x ) )
\mathbf{f}: X \to \mathbb{R}^{k} \quad \text{and} \quad \mathbf{f}(x)=(f_{1}(x),\cdots,f_{k}(x))
f : X → R k and f ( x ) = ( f 1 ( x ) , ⋯ , f k ( x ))
Then
(a) A necessary and sufficient condition for f \mathbf{f} f to be continuous is that each of the f 1 , ⋯ , f k f_{1},\cdots,f_{k} f 1 , ⋯ , f k is continuous.
(b) Suppose g \mathbf{g} g is a function defined in the same way as f \mathbf{f} f . If f \mathbf{f} f , g \mathbf{g} g are continuous, then f + g \mathbf{f}+\mathbf{g} f + g , f ⋅ g \mathbf{f}\cdot \mathbf{g} f ⋅ g are also continuous.
Proof (a) Assuming each of f i f_{i} f i is continuous, by the definition of continuity, for each of f i , ε i f_{i}, \varepsilon_{i} f i , ε i there exists
d X ( x , p ) < δ i ⟹ ∣ f i ( x ) − f i ( p ) ∣ < ε i ( i = 1 , ⋯ , k )
d_{X}(x,p)<\delta_{i} \implies \left| f_{i}(x)-f_{i}(p) \right|< \varepsilon_{i} \quad(i=1,\cdots,k)
d X ( x , p ) < δ i ⟹ ∣ f i ( x ) − f i ( p ) ∣ < ε i ( i = 1 , ⋯ , k )
such that
d X ( x , p ) < δ ⟹ ∣ f ( x ) − f ( p ) ∣ = ∑ i = 1 k ∣ f i ( x ) − f i ( p ) ∣ 2 < ∑ i = 1 k ε i = ε
d_{X}(x,p)<\delta \implies \left| \mathbf{f}(x)-\mathbf{f}(p) \right|=\sqrt{\sum\limits_{i=1}^{k}\left|f_{i}(x) -f_{i}(p) \right|^{2}}<\sum \limits _{i=1} ^{k}\varepsilon_{i}=\varepsilon
d X ( x , p ) < δ ⟹ ∣ f ( x ) − f ( p ) ∣ = i = 1 ∑ k ∣ f i ( x ) − f i ( p ) ∣ 2 < i = 1 ∑ k ε i = ε
Therefore, f \mathbf{f} f is continuous. Conversely, if we assume f \mathbf{f} f is continuous, then for some ε > 0 \varepsilon>0 ε > 0
d X ( x , p ) < δ ⟹ ∣ f i ( x ) − f i ( p ) ∣ ≤ ∣ f ( x ) − f ( p ) ∣ < ε
d_{X}(x,p)<\delta \implies \left|f_{i}(x)-f_{i}(p) \right|\le \left|\mathbf{f}(x)-\mathbf{f}(p) \right|<\varepsilon
d X ( x , p ) < δ ⟹ ∣ f i ( x ) − f i ( p ) ∣ ≤ ∣ f ( x ) − f ( p ) ∣ < ε
Thus, each of f i f_{i} f i is continuous.
■
(b) It holds by Theorem 1 and (a) .
■