Derivative of a function of another function

  • Context: Undergrad 
  • Thread starter Thread starter kent davidge
  • Start date Start date
  • Tags Tags
    Derivative Function
Click For Summary

Discussion Overview

The discussion revolves around the differentiation of a function that depends on another function, specifically focusing on the interpretation of derivatives in the context of the chain rule. Participants explore notation issues and the implications of variable dependencies in mathematical expressions.

Discussion Character

  • Exploratory
  • Technical explanation
  • Debate/contested
  • Mathematical reasoning

Main Points Raised

  • One participant questions how to view the derivative ##\frac{\partial \vartheta'}{\partial \varphi}## given that ##\vartheta'(\vartheta) = \vartheta## and ##\vartheta = \vartheta(\varphi)##, noting that ##\vartheta(\varphi)## is not linear in ##\varphi##.
  • Another participant seeks clarification on the meaning of ##\vartheta'(\vartheta)##, suggesting that the notation may be misleading.
  • There is a consensus among some participants that the notation is problematic, particularly because the prime notation typically indicates a derivative.
  • Some participants propose that the discussion relates to the chain rule, with one participant expressing uncertainty about how to proceed with the derivative when ##\varphi## is not the direct argument of the function.
  • A later reply confirms the use of the chain rule with partial derivatives, presenting the formula ##\frac{\partial f}{\partial \varphi} = \frac{\partial f}{\partial g}\frac{\partial g}{\partial \varphi}##.
  • One participant provides an example involving functions ##f(y) = 2y## and ##y = x^2##, questioning the interpretation of ##\frac{df}{dx}## and discussing the implications of variable domains.
  • Another participant critiques the use of Greek letters in the notation, suggesting that clearer examples would improve understanding.

Areas of Agreement / Disagreement

Participants express disagreement regarding the notation and its implications for understanding derivatives. There is no consensus on the best way to represent the functions or the derivatives involved, and multiple interpretations of the mathematical expressions remain.

Contextual Notes

Participants highlight limitations in notation and the potential for ambiguity when discussing functions with different domains. The discussion reflects a range of interpretations regarding the application of the chain rule and the treatment of derivatives in this context.

kent davidge
Messages
931
Reaction score
56
This is really a simple question, but I'm stuck.

Suppose we have a function ##\vartheta'(\vartheta) = \vartheta## and that ##\vartheta = \vartheta(\varphi)## and we know what ##\vartheta(\varphi)## is. How should I view ##\frac{\partial \vartheta'}{\partial \varphi}##? Should I set it equal to zero because ##\varphi## does not appear explicitely? But as I said, I know what ##\vartheta(\varphi)## is in this particular case, and it is not even linear in ##\varphi##, so it doesn't make sense to say that its derivatives vanish.

Edit: I added a crucial correction to my post.
 
Physics news on Phys.org
What does ## \vartheta '( \vartheta )## mean?
 
fresh_42 said:
What does ## \vartheta '( \vartheta )## mean?
I'm sorry, that was a bad choice of notation. ##\vartheta'## is simply a function, the prime is there to differentiate it from ##\vartheta##. It does not mean derivative. If you prefer, let's call it ##\Theta## from now on to cause no confusion.
 
Still truly bad notation, especially because a prime superscript placed on the symbol for a function typically means its derivative. And this question is about derivatives.
 
  • Like
Likes   Reactions: fresh_42
zinq said:
Still truly bad notation, especially because a prime superscript placed on the symbol for a function typically means its derivative. And this question is about derivatives.
:oops::cry:
 
So we are talking about the chain rule? ##g=g(\varphi)## and ##f=f(g)##, so ##f=f(g(\varphi))##?
 
fresh_42 said:
So we are talking about the chain rule? ##g=g(\varphi)## and ##f=f(g)##, so ##f=f(g(\varphi))##?
oh yes. And I would like to know $$\frac{\partial f}{\partial \varphi}$$ If ##\varphi## were the argument of ##f## directly, then it would be the same as the total derivative of ##f## wrt it. But since ##\varphi## is the argument of ##g##, I'm not sure how to proceed.

chain rule?
oops, this just reminded me that we can use the chain rule also with partial derivatives. so $$\frac{\partial f}{\partial \varphi} = \frac{\partial f}{\partial g}\frac{\partial g}{\partial \varphi}$$ :biggrin::biggrin::biggrin:
 
##\dfrac{df}{d \varphi}=\dfrac{df(g)}{dg}\cdot\dfrac{dg(\varphi)}{d\varphi}##.
 
Let's try different notation for the OP and use an example.

Suppose we are given ##f(y) = 2y## and ## y = x^2##. Then what is ##\frac {df}{dx}##?

Is it zero because ##x## does not appear in the formula ##f(y) = 2y## ? Or is it ##4x## because ##f(x) = 2(x^2)## ?

This involves ambiguous and bad notation - and notation that is customary, especially in writing about physics! Beginning students in mathematics are taught that a function has a definite domain and co-domain. So if two functions have different co-domains, they must be different functions. The function ##f(y)## can take on negative values. The function ##f(x)## cannot. So we shouldn't use the name ##f## for both functions. However, it is common practice to do so.

When we are "given" that ##y = x^2##, it isn't clear how the name ##f## should be used in representing this fact. To be unambiguous, we could say there is function ##f(y) = 2y## and another function ##B(x) = f(x^2) = 2x^2##.

Technically speaking, if ##x## denotes a variable not in the domain of ##f## then ##\frac{df}{dx}## is undefined. However, convention says that we interpret ##\frac{df}{dx}## by pretending ##f(y)## is actually a function of two variables ##(x,y)## that is constant with respect to ##x##. With that convention, ##\frac{df}{dx} = \frac{\partial f}{\partial_x} = 0##.

I'd say the most common interpretation of the above example in physics is that ##\frac{df}{dx} = 0##.

In a math text, the author might want to give students an exercise in the chain rule. He would expect them to interpret ##\frac{df}{dx}## as the derivative of the function ##B(x)##.
 
  • Like
Likes   Reactions: kent davidge
  • #10
kent davidge said:
Suppose we have a function ##\vartheta'(\vartheta) = \vartheta## and that ##\vartheta = \vartheta(\varphi)## and we know what ##\vartheta(\varphi)## is.
Why are you using Greek letters here? The examples shown in the posts by @fresh_42 and @Stephen Tashi are much clearer, not to mention easier to type.
 
  • Like
Likes   Reactions: kent davidge
  • #11
Mark44 said:
Why are you using Greek letters here? The examples shown in the posts by @fresh_42 and @Stephen Tashi are much clearer, not to mention easier to type.
Indeed... As if it wasn't enough not understanding the topic, I come up with this notation to make things more difficult.
 

Similar threads

  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 16 ·
Replies
16
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 6 ·
Replies
6
Views
3K
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
1K