# Homework Help: Numerical methods

1. Sep 28, 2005

### stunner5000pt

Find intervals containing solutions to
$$4x^2 - e^x = 0$$
well someone suggested i sketch the graphs of 4x^2 and 2^x but im not sure on how to go past that point...
All i have to do is find the intervals so do the intersection point(s) of these two functions indicate the intervals where the roots may exist?

Find the Rates of convergence of the following sequences are n -> infinity

a) $$\lim_{n\rightarrow\infty} \sin{\frac{1}{n}} = 0$$
i was wondering if expanding the sine by taylor series would help the case.. and if that is so, at which term would i stop at?

b) $$\lim_{n\rightarrow\infty} \sin{\frac{1}{n^2}} = 0$$
c) $$\lim_{n\rightarrow\infty} (\sin{\frac{1}{n}})^2 = 0$$
d) $$\lim_{n\rightarrow\infty} \ln{(n+1)}-\ln{n}= 0$$

the last one looks like ln((n+1)/n) which i think can be expanded by taylor series... if this is the right idea please do tell.
ALso what is the taylor series about.. as in what is the x0 value and the c value where c is the n+1 derivative's evaltuation point.

any help that you could provide for these would be greatly appreicated!!

2. Sep 28, 2005

### Integral

Staff Emeritus
Intersections of the 2 separate functions will BE roots of the given function. Any interval containing an intersection will also contain a root. You can also evaluate the function looking for sign changes, any interval, say (a,b), for which f(a)f(b)<0 holds, contains a root or odd number of roots.

3. Sep 28, 2005

### stunner5000pt

im not quite sure how to apply that here...
i mena 4x^2 and e^x are always positive so the f(a)f(b) is never zero

and to pick the intervals of intersection... isnt that just guessing?

4. Sep 28, 2005

### hotvette

Hint: add $$e^x$$ to both sides of the equation.

5. Sep 28, 2005

### stunner5000pt

how would adding e^x to both sides help in any way?? What im trying to do is find the intervals where 4x^2 = e^x..

6. Sep 28, 2005

### hotvette

Just to illustrate that what you are looking for is one equal to the other, meaning where the graphs intersect. Or, did I miss something? Looks like you already got that far. So, what do you need, the exact x?

Last edited: Sep 28, 2005
7. Sep 28, 2005

### stunner5000pt

well im trying to find out the intervals and not the points themselves of where the two graphs intersect.
so like integral said i have to pick some interval where these intersections (i think there are two) lie.

8. Sep 28, 2005

### calvino

hmmm

well... Firstly, take the limits of the funtion as they approach -infinity and +infinity. This should show you the behaviour of the graphs as the go "inward". [ Nextly, try dividing the f'n in half to see how it further behaves. (i'm not sure if this will help you much) ]. Furthermore, ever heard of ivt?

[ever try and find solutions to a quintic?]-CALVIN

9. Sep 28, 2005

### calvino

hint: I think there are more than 2, but that's just me.

10. Sep 28, 2005

### stunner5000pt

im not usre how i can use that ...
yes i have heard of IVT
but is finding the two points just a matter of gueswork? If that were the case then it would be daunting if there were more than two solutions...

11. Sep 28, 2005

### calvino

Well, if you do the first part I mentioned, and see how the function behaves outwards, you've done a good part of the work. This leads to IVT, which should be the easy part. It may seem like guesswork (and it is), but who says guesswork is bad? -Calvin

12. Sep 28, 2005

### calvino

Oh, and having more than 2 solutions is not "daunting". (Thinking Poly's?)

13. Sep 28, 2005

### stunner5000pt

ok for the convergence of the limits - is the rate of convergence for the first limit 1/n^2, for the second limit (1/n^4)?