Second Derivative of a Composition of Functions

skyline01
Messages
3
Reaction score
0

Homework Statement


Derive an expression for the composition of 2 functions.


Homework Equations





The Attempt at a Solution


I started with supposing that y(x) = f(g(x)). I know that dy/dx = df/dg * dg/dx (via the chain rule). Doing the derivative again, I started with the product rule:
d2y/dx2 = d/dx (df/dg * dg/dx) = df/dg * d2g/dx2 + dg/dx * d/dx (df/dg).

I don't know what d/dx (df/dg) means. I think I'm supposed to do the chain rule again on this part, but I'm not totally clear. Thanks!
 
Physics news on Phys.org
skyline01 said:

Homework Statement


Derive an expression for the composition of 2 functions.

Homework Equations



The Attempt at a Solution


I started with supposing that y(x) = f(g(x)). I know that dy/dx = df/dg * dg/dx (via the chain rule). Doing the derivative again, I started with the product rule:
d2y/dx2 = d/dx (df/dg * dg/dx) = df/dg * d2g/dx2 + dg/dx * d/dx (df/dg).

I don't know what d/dx (df/dg) means. I think I'm supposed to do the chain rule again on this part, but I'm not totally clear. Thanks!
Hello skyline01. Welcome to PF !

One suggestion is to to look at some specific examples, such as y(x) = sin(x5).

Then f(x) = sin(x), g(x) = x5 .
 
Thank you, SammyS! I tried the sample function you suggested, but I'm still not sure if I am expressing d/dx (df/dg) correctly. Here is how your sample function worked out:

y(x) = sin(x5)
So, f(x) = sin(x) and g(x) = x5.
Therefore,
dy/dx = cos(x5) 5x4
and
d2y/dx2 = cos(x5) 20x3 + 5x4 (-sin(x5)) 5x4.

Generalizing this to any composition of 2 functions, we have
d2y/dx2 = df/dg * d2g/dx2 + dg/dx * d2f/dg2.

Does this look right?
 
skyline01 said:
Thank you, SammyS! I tried the sample function you suggested, but I'm still not sure if I am expressing d/dx (df/dg) correctly. Here is how your sample function worked out:

y(x) = sin(x5)
So, f(x) = sin(x) and g(x) = x5.
Therefore,
dy/dx = cos(x5) 5x4
and
d2y/dx2 = cos(x5) 20x3 + 5x4 (-sin(x5)) 5x4.

Generalizing this to any composition of 2 functions, we have
d2y/dx2 = df/dg * d2g/dx2 + dg/dx * d2f/dg2.

Does this look right?
Almost correct.

d2y/dx2 = df/dg * d2g/dx2 + (dg/dx)2 * d2f/dg2.
 
Yes, you are correct. Thank you for catching that. So,
d/dx(df/dg) = dg/dx * d2f/dg2.

I find it strange that in every calculus book I have ever read (including real analysis books), they always stop at the first derivative when discussing the chain rule.
 
Yes, that is strange.

The notation can be confusing, even for 1st derivatives, but as you mentioned in your initial post, it gets quite a bit more confusing with higher order derivatives. That's why I suggested differentiating a concrete example.

Here's a link from Wikipedia: http://en.wikipedia.org/wiki/Chain_rule#Higher_derivatives.

Have a good one !
SammyS
 
Prove $$\int\limits_0^{\sqrt2/4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx = \frac{\pi^2}{8}.$$ Let $$I = \int\limits_0^{\sqrt 2 / 4}\frac{1}{\sqrt{x-x^2}}\arcsin\sqrt{\frac{(x-1)\left(x-1+x\sqrt{9-16x}\right)}{1-2x}} \, \mathrm dx. \tag{1}$$ The representation integral of ##\arcsin## is $$\arcsin u = \int\limits_{0}^{1} \frac{\mathrm dt}{\sqrt{1-t^2}}, \qquad 0 \leqslant u \leqslant 1.$$ Plugging identity above into ##(1)## with ##u...
Back
Top