Prove that a product of continuous functions is continuous

Click For Summary

Homework Help Overview

The discussion revolves around proving that the product of continuous functions is continuous, specifically in the context of complex functions. Participants explore the implications of continuity definitions and the application of the epsilon-delta method in this proof.

Discussion Character

  • Exploratory, Conceptual clarification, Mathematical reasoning

Approaches and Questions Raised

  • Some participants attempt to establish continuity through direct application of definitions and inequalities, while others question the validity of certain steps and the assumptions made regarding the functions involved. There is also discussion about the implications of different domains for the functions.

Discussion Status

The discussion is ongoing, with participants providing insights and outlining potential approaches. Some have noted the need for careful justification of each step in the proof, while others have expressed uncertainty about specific logical connections. There is no explicit consensus yet, but various lines of reasoning are being explored.

Contextual Notes

Participants mention the importance of addressing technicalities, such as the behavior of functions at specific points and the implications of continuity when functions have different domains. There is also a reference to external resources for further clarification.

docnet
Messages
796
Reaction score
486
Homework Statement
.
Relevant Equations
complex analysis
Screen Shot 2021-12-23 at 1.53.25 PM.png


##f## is continuou on ##\mathbb{C}##, so for al ##\epsilon>0##, there is a ##\delta>0## such that $$|\tilde{z}-z|\leq \delta \Rightarrow |f(\tilde{z})-f(z)|\leq \epsilon$$ for all ##\tilde{z}## and ##z## in ##\mathbb{C}##.

Complex conjugation is a norm preserving operation on ##\mathbb{C}##, so

$$|f(\tilde{z})-f(z)|=|\bar{f}(\tilde{z})-\bar{f}(z)|\leq \epsilon$$

The same ##\delta## satisfies the definition of continuity for ##\bar{f}##, so ##\bar{f}## is continuous on ##\mathbb{C}##.

The product of two continuous functions is continuous, so ##g=f\cdot \bar{f}## is continuous.

The conclusion feels shaky or like I'm not doing the real work of the problem somehow.
 
Last edited:
Physics news on Phys.org
docnet said:
The absolute value function is a norm preserving operation on ##\mathbb{C}##,
I think you mean complex conjugation is norm preserving.

docnet said:
The product of two continuous functions is continuous, so ##g=f\cdot \bar{f}## is continuous.
If you are allowed to invoke that theorem, then this is fine.
 
  • Love
Likes   Reactions: docnet
PeroK said:
I think you mean complex conjugation is norm preserving.
Yes! I edited my post to fix my mistake.
 
Not using the theorem:

$$|\bar{f}(\tilde{z})f(\tilde{z})-\bar{f}(z)f(z)|=|\bar{f}(\tilde{z})f(\tilde{z})-\bar{f}(z)f(z)+\bar{f}(\tilde{z})f(z)-\bar{f}(\tilde{z})f(z)|$$
$$\leq |\bar{f}(\tilde{z})||f(\tilde{z})-f(z)|+|f(z)||\bar{f}(\tilde{z})-\bar{f}(z)|= |\bar{f}(\tilde{z})|\epsilon +|f(z)|\epsilon$$Let ##\delta=\frac{\epsilon }{|\bar{f}(\tilde{z})|+|f(z)|}## then ##\bar{f}f## is continuous by the epsilon delta definition of continuity.
 
Last edited:
docnet said:
Thank you, I'll look into this and come up with a solution soon.
It's trickier than it looks. If you really want to master the epslion-delta technique then you have to really knuckle down and battle through something like this. Every step has to be justified and every nuance covered. For example, what happens to your proof if ##f(z) = f(\tilde z) = 0##?

The best plan, I find, is to get the strategy/outline first before diving in with the technical details.

Let me post an outline to show you what I mean ...
 
  • Love
Likes   Reactions: docnet
Outline proof (I'll do it for real ##f, g## just to simplify the notation a bit):

1) The first thing we must do is state clearly what we are going to prove:

Let ##f, g## be continuous at some point ##a##. Then the product function ##fg## is continuous at ##a##.

2) First technicality: what happens if ##f## and ##g## have different domains? What if ##f(x) = \sqrt x##, with domain ##[0, \infty)##; and, ##g(x) = \sqrt {-x}##, with domain ##(-\infty, 0]##. The product function is only defined at ##x = 0##. That's okay, because then ##fg## is continuous vacuously at ##x = 0##.

3) Do we mention domains further or just gloss over that point? For simplicity, let's assume that we have restricted our attention to a shared domain that inlcudes ##a## and not say any more about this.

4) The key inequality is: $$|f(x)g(x) - f(a)g(a)| \le |f(x)||g(x) - g(a)| + |g(a)||f(x) - f(a)|$$How do we deal with this?

4a) Note that ##|g(a)|## is some fixed number (but it might equal ##0##, so we need to be careful about that). The term ##|f(x) - f(a)|## can be made arbitrarily small. We're okay there.

4b) We can make ##|f(x)|## close to ##|f(a)|##, so we must be able to limit the size of this term. And, the term ##|g(x) - g(a)|## can be made arbitrarily small. We're okay here as well.

4c) We have the sum of two expressions, so we should try to make each ##< \frac \epsilon 2##.

5) Let's look at each of those points in more detail.

5a) If ##|g(a)| \le 1##, then the first term is easy. We can find ##\delta_1## such that ##|f(x) - f(a)| < \frac \epsilon 2##. And, if ##|g(a)| > 1##, then we need to find ##\delta_2## such that ##|f(x) - f(a)| < \frac \epsilon {2|g(a)|}##.

Alternatively, if ##|g(a)| = 0##, then the first term vanishes and otherwise we can find ##\delta_2## such that ##|f(x) - f(a)| < \frac \epsilon {2|g(a)|}##.

Another alternative, we could find ##\delta_3## such that ##|f(x) - f(a)| < \frac \epsilon {2(1+ |g(a)|)}##. That takes care of all cases.

We need to decide on one of those approaches. Pick whatever looks best.

5b) The term ##|f(x)|## is trickier, because ##x## is a variable. But, we can find ##\delta_4## such that ##|f(x)| < |f(a)| + 1##. We should be able to show that from the triangle inequality.

That just leaves ##\delta_5## such that ##|g(x) - g(a)| < \frac \epsilon{2(|f(a)| + 1)}##.

5c) We just need to decide on the best approach, rationalise our deltas, and take ##\delta = min(\delta_1, \delta_2)## and we are good to go.

6) How do all epsilon-delta proofs start?

Let ##\epsilon > 0 \dots##
 
Last edited:
  • Love
Likes   Reactions: docnet
PeroK said:
Outline proof (I'll do it for real ##f, g## just to simplify the notation a bit):

1) The first thing we must do is state clearly what we are going to prove:

Let ##f, g## be continuous at some point ##a##. Then the product function ##fg## is continuous at ##a##.

2) First technicality: what happens if ##f## and ##g## have different domains? What if ##f(x) = \sqrt x##, with domain ##[0, \infty)##; and, ##g(x) = \sqrt {-x}##, with domain ##(-\infty, 0]##. The product function is only defined at ##x = 0##. That's okay, because then ##fg## is continuous vacuously at ##x = 0##.

3) Do we mention domains further or just gloss over that point? For simplicity, let's assume that we have restricted our attention to a shared domain that inlcudes ##a## and not say any more about this.

4) The key inequality is: $$|f(x)g(x) - f(a)g(a)| \le |f(x)||g(x) - g(a)| + |g(a)||f(x) - f(a)|$$How do we deal with this?

4a) Note that ##|g(a)|## is some fixed number (but it might equal ##0##, so we need to be careful about that). The term ##|f(x) - f(a)|## can be made arbitrarily small. We're okay there.

4b) We can make ##|f(x)|## close to ##|f(a)|##, so we must be able to limit the size of this term. And, the term ##|g(x) - g(a)|## can be made arbitrarily small. We're okay here as well.

4c) We have the sum of two expressions, so we should try to make each ##< \frac \epsilon 2##.

5) Let's look at each of those points in more detail.

5a) If ##|g(a)| \le 1##, then the first term is easy. We can find ##\delta_1## such that ##|f(x) - f(a)| < \frac \epsilon 2##. And, if ##|g(a)| > 1##, then we need to find ##\delta_2## such that ##|f(x) - f(a)| < \frac \epsilon {2|g(a)|}##.

Alternatively, if ##|g(a)| = 0##, then the first term vanishes and otherwise we can find ##\delta_2## such that ##|f(x) - f(a)| < \frac \epsilon {2|g(a)|}##.

Another alternative, we could find ##\delta_3## such that ##|f(x) - f(a)| < \frac \epsilon {2(1+ |g(a)|)}##. That takes care of all cases.

We need to decide on one of those approaches. Pick whatever looks best.

5b) The term ##|f(x)|## is trickier, because ##x## is a variable. But, we can find ##\delta_4## such that ##|f(x)| < |f(a)| + 1##. We should be able to show that from the triangle inequality.

That just leaves ##\delta_5## such that ##|g(x) - g(a)| < \frac \epsilon{2(|f(a)| + 1)}##.

5c) We just need to decide on the best approach, rationalise our deltas, and take ##\delta = min(\delta_1, \delta_2)## and we are good to go.

6) How do all epsilon-delta proofs start?

Let ##\epsilon > 0 \dots##
please allow me plenty of time to digest this.. and Merry Christmas!
 
  • Like
Likes   Reactions: PeroK

Similar threads

  • · Replies 20 ·
Replies
20
Views
3K
  • · Replies 6 ·
Replies
6
Views
2K
Replies
8
Views
2K
  • · Replies 14 ·
Replies
14
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K
Replies
7
Views
1K
Replies
4
Views
2K
  • · Replies 11 ·
Replies
11
Views
2K
  • · Replies 26 ·
Replies
26
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K