Chain rule and change of variables again

Join the discussion
Registration is free. Start your own thread to ask a follow-up.
3 replies · 2K views
jonjacson
Messages
450
Reaction score
38
We start with:

d2y/dx2

And we want to consider x as function of y instead of y as function of x.

I understand this equality:

dy/dx = 1/ (dx/dy)

But for the second order this equality is provided:

d2y/dx2 =- d2x/dy2 / (dx/dy)3

Does anybody understand where is it coming from? The cubic term in the denominator looks quite strange, I don't know how to understand that equation.

ANy help is welcome.
 
Physics news on Phys.org
jonjacson said:
We start with:

d2y/dx2

And we want to consider x as function of y instead of y as function of x.

I understand this equality:

dy/dx = 1/ (dx/dy)

But for the second order this equality is provided:

d2y/dx2 =- d2x/dy2 / (dx/dy)3

Does anybody understand where is it coming from? The cubic term in the denominator looks quite strange, I don't know how to understand that equation.

ANy help is welcome.
Write your second equation above as ##dy/dx = (dx/dy)^{-1}##
Now take the derivative with respect to x of both sides:
##d^2y/dx^2 = (-1) (dx/dy)^{-2} \cdot d/dx(dx/dy)##
On the right side above, I'm using the chain rule.
Is that enough of a start?
 
Mark44 said:
Write your second equation above as ##dy/dx = (dx/dy)^{-1}##
Now take the derivative with respect to x of both sides:
##d^2y/dx^2 = (-1) (dx/dy)^{-2} \cdot d/dx(dx/dy)##
On the right side above, I'm using the chain rule.
Is that enough of a start?

It looks like a great start but I have a question.

If we have x as function of y, dx/dy will be a function of y too, WHat is the meaning of deriving this respect to x?
X is the independent variable now, Does it make sense to differenciate a function respect itself?

Maybe I am confused and I should go to bed.
 
jonjacson said:
It looks like a great start but I have a question.

If we have x as function of y, dx/dy will be a function of y too, WHat is the meaning of deriving this respect to x?
You can differentiate it with respect to x (not derive it).
jonjacson said:
X is the independent variable now
No, if x is a function of y, x is the dependent variable, and y is the independent variable.
jonjacson said:
, Does it make sense to differenciate a function respect itself?

Maybe I am confused and I should go to bed.
In the last part of what I wrote I have ##d/dx(dx/dy)##. That's the same as ##\frac d {dy} \left(\frac {dx}{dy} \right)\cdot \frac{dy}{dx}##, with the chain rule being applied once more.