What is the derivative of a modulus function?

  • Thread starter Thread starter Zurtex
  • Start date Start date
  • Tags Tags
    Modulus
Zurtex
Science Advisor
Homework Helper
Messages
1,118
Reaction score
1
Feeling a little bit more confident about my calculus skills I was hoping to check if this is correct. Let’s say you have:

g(|f(x)|)

And you want to take the derivative with respect to x, well using the chain rule you get:

\bigl( g(|f(x)|) \bigr)' = (|f(x)|)' g'(|f(x)|)

Looking more closly at (|f(x)|)'. Defining it as:

\left( +\sqrt{(f(x))^2} \right)'

Then using the chain rule we get:

(|f(x)|)' = 2f(x) f'(x) \: \frac{1}{2} \: \frac{1}{ +\sqrt{(f(x))^2} }

Simplifying:

(|f(x)|)' = f'(x) \frac{f(x)}{|f(x)|}

Substituting back in the original equation and sorting out the problem of when f(x) = 0:

\bigl( g(|f(x)|) \bigr)&#039; = \begin{cases}<br /> f&#039;(x) \frac{f(x)}{|f(x)|}g&#039;(|f(x)|) &amp; \text{if $ f(x) \neq 0$} \\ \\<br /> \lim_{f(x) \rightarrow 0} \left(f&#039;(x) \frac{f(x)}{|f(x)|}g&#039;(|f(x)|) \right) &amp; \text{if $ f(x) = 0$ and $ \lim_{f(x) \uparrow 0} \left( f&#039;(x) \frac{f(x)}{|f(x)|}g&#039;(|f(x)|) \right) = \lim_{f(x) \downarrow 0} \left( f&#039;(x) \frac{f(x)}{|f(x)|}g&#039;(|f(x)|) \right)$} \\ \\<br /> \text{Undefinied} &amp; \text{if $ f(x) = 0$ and $ \lim_{f(x) \uparrow 0} \left( f&#039;(x) \frac{f(x)}{|f(x)|}g&#039;(|f(x)|) \right) \neq \lim_{f(x) \downarrow 0} \left( f&#039;(x) \frac{f(x)}{|f(x)|}g&#039;(|f(x)|) \right)$}<br /> \end{cases}<br />

Correct? I know it all seems a bit over the top (especially when you look at the tex for it :wink:) but I like things to be well defined and in a form like this where I understand it better.

Edit: I made a mistake, it should be correct now :smile:
 
Last edited:
Physics news on Phys.org
it is over the top. it is easier to use mod of Y is Y if Y is positive and -Y if Y is negative, the derivative not defined if Y=0, though it is conceivalbe that g(modY) is differentiable when Y=0.
 
matt grime said:
it is over the top. it is easier to use mod of Y is Y if Y is positive and -Y if Y is negative, the derivative not defined if Y=0, though it is conceivalbe that g(modY) is differentiable when Y=0.
I'm not trying to find an easier method, but rather checking that my method is correct.

Although I do not claim this method to be more useful. I think it does yeld a more accurate result over some functions than the method you use, for example if you want to find the derivative with respect of x:

\left| x^n \right| \quad n \in \mathbb{N} \backslash \negmedspace \{ 1 \}

But really more than anything I just want to check it is correct.

Edit: I just realized a big mistake in it, but I've now edited the first post accordingly.
 
Last edited:
but they are easily seen to be equivalent, indeed your f/modf is simply +/-1 depending on whether f is positive or negative, ie you are just not using the words 'when something is positive/negative' and making it neater, so yes your method is correct in spirit though i haven't checked the detail
 
matt grime said:
but they are easily seen to be equivalent, indeed your f/modf is simply +/-1 depending on whether f is positive or negative, ie you are just not using the words 'when something is positive/negative' and making it neater, so yes your method is correct in spirit though i haven't checked the detail
Neither do I write:

The derivative of the positive square root of a squared function of x, instead I represent it like:

\left( +\sqrt{(f(x))^2} \right)&#039;

It just fits easier in my head if I have an algebraic notation for something rather than a word notation (probably something to do with my dyslexia but whatever). Anyway, I'm glad the spirit is right, but I think I lack the correct notation with the limits. It is meant to be x approaching from smaller values (values coming from negative infinity) to reach f(x) = 0, but instead I think I've wrote f(x) approaching from smaller values to f(x) = 0, which may not make too much sense for functions like f(x) = x2.
 
Back
Top