# Hey, I can't solve this limit:'(

1. Jun 7, 2005

### schattenjaeger

Assuming it exists(and upon checking the answer is 1, but I'll be damned if I can get it)

(e^x * tan^-1 y) / y

the limit as (x,y)->(0,0)

I can't find a path where the denominator isn't 0 or cancels out that goes through the point 0,0, so I can't, you know, solve it.

2. Jun 7, 2005

### Berislav

What is the Gauss aproximation for tan^-1 for small angles? e^0=1. Not the most exact way, but it should work.

3. Jun 7, 2005

### schattenjaeger

Well, I'm in Calculus 3 and have no clue what that approxamation is nor do I think I'm expected to for this problem, thanks though

4. Jun 7, 2005

### vsage

See next post.

Last edited by a moderator: Jun 7, 2005
5. Jun 7, 2005

### schattenjaeger

Well, it's a function of two variables, and this problem is in the chapter before partial derivatives, so ignoring the fact that I'm not technically sposed to know how to apply l'hopital's rule to it, I don't even know if you can on a multivariable thing like that

6. Jun 7, 2005

### vsage

Alright that puts the question in a little perspective. Still, you can simplify the limit to one variable by just applying x = 0 to it.

Edit: Applying tan^-1() to each side of the first approximation Berislav gave would work also I guess.

Last edited by a moderator: Jun 7, 2005
7. Jun 7, 2005

### Berislav

Gauss aproximation:

siny=y, for very small y --->sin^-1y=y

cosy=sqrt(1-y^2)

tany=y/(sqrt(1-y^2))

The y's will cancel when you combine the above eqaution with your own.

Last edited: Jun 7, 2005
8. Jun 7, 2005

### schattenjaeger

Then I have to take the derivative of tan^-1y, which definately isn't 1(like it would have to be for l'hopital's rule to give me 1)

As for the guass approximation, I don't think I should use it here because it doesn't specify only small values of y, and I certainly haven't officially learned it yet

9. Jun 7, 2005

### Hurkyl

Staff Emeritus
For small values of y, sin y = y, cos y = 1, and tan y = y. :tongue2:

10. Jun 7, 2005

### vsage

How's it not? $$\frac{d(tan^{-1}(y))}{dy} = \frac{1}{1+y^2}$$ at y = 0.. Divide this by the derivative of the denominator of your original limit.

11. Jun 7, 2005

### Berislav

Right. I only approximated sin y=y, though.

12. Jun 7, 2005

Oh right

ok, thanks!