# Z=f(x,y), x=function, y=function. dz/dx=

1. Apr 24, 2013

### mrcleanhands

1. The problem statement, all variables and given/known data
Consider z=f(x,y), where x=rcosθ and y=rsinθ

Show that $\frac{\partial z}{\partial x}=\cos\theta\frac{\partial z}{\partial r}-\frac{1}{r}\sin\theta\frac{\partial z}{\partial\theta}$

2. Relevant equations

3. The attempt at a solution

Z
Connects to X and Y
X Connects to r and θ, Y Connects to r and θ

Is dz/dx not then equal to dx/dr + dx/dθ? because that is equivalent to dz/dx = r -rsinθ.
What am I doing wrong here?

2. Apr 24, 2013

### Simon Bridge

Evaluate $\frac{\partial z}{\partial r}$

3. Apr 24, 2013

### mrcleanhands

$\frac{\partial z}{\partial r}=\cos\theta\frac{\partial z}{\partial x}+\sin\theta\frac{\partial z}{\partial y}$

then rearranging this $\frac{\partial z}{\partial x}=\sec\theta\frac{\partial z}{\partial r}-\tan\theta\frac{\partial z}{\partial y}$

I'm not doing something right or not understanding some underlying concept here. Why was my dz/dx derivation wrong?

4. Apr 24, 2013

### Simon Bridge

Don't rearrange yet. Do the other partial - the one with theta.

5. Apr 25, 2013

### mrcleanhands

$\frac{\partial z}{\partial\theta}=-r\sin\theta\frac{\partial z}{\partial x}+\frac{1}{r}\cos\theta\frac{\partial z}{\partial y}$

I've re-arranged this and get $\frac{\partial z}{\partial x}=-\frac{1}{r}\csc\frac{\partial z}{\partial\theta}+\frac{1}{r^{2}}\cot\theta\frac{\partial z}{\partial y}$

which should be the same as $\frac{\partial z}{\partial x}=\sec\theta\frac{\partial z}{\partial r}-\tan\theta\frac{\partial z}{\partial y}$

Now I'm more confused...

Isn't it just $r-r\sin\theta$?

6. Apr 25, 2013

### Simon Bridge

Start with the end in mind or you won't get anywhere:
- you want ∂z/∂x in terms of ∂z/∂r and ∂z/dθ.

You don't need any terms in ∂z/∂y ... but the above relations each have such a term.
This should suggest a course of action to you.

7. Apr 26, 2013

### mrcleanhands

Hey Simon, I got it!

I isolated $\frac{\partial z}{\partial y}$ for both equations of $\frac{\partial z}{\partial r}$ and $\frac{\partial z}{\partial \theta}$ and then equated the two and re-arranged a little to get $\frac{\partial z}{\partial x}=\cos\theta\frac{\partial z}{\partial r}-\frac{1}{r}\sin\theta\frac{\partial z}{\partial\theta}$.

However, I still don't get why the solution is simply not $\frac{\partial z}{\partial x}= \frac{\partial x}{\partial r}+ \frac{\partial x}{\partial \theta}=\cos\theta - r\sin\theta$

8. Apr 27, 2013

### Simon Bridge

Well done.

... but didn't you just demonstrate that it isn't?
Looks like you may need to go back to what the partial means.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted