The Student Room Group

Analysis (Tripos Question) Help

http://www.maths.cam.ac.uk/undergrad/pastpapers/2005/Part_IA/PaperIA_1.pdf

Hi, I need help on 9F) iii) and the last part of 11E)

For 9F) iii) I tried the ratio test, I can't think of any other test which would work and I think I for the last part of 11 E) I need an auxiliary function, I tried f(x) = g(x + 1/n ) - g(x) , it doesn't quite give me the result.

Thanks for any help in advance.
Reply 1
For 9F (iii): You should be able to see the terms behave like 1n\dfrac{1}{n} for large n. So you're expecting it to diverge. Since you're expecting it to diverge, the way forwards is to show that for large n the numerator is > kn^3 for some constant k, and the denominator is < Kn^4 for some other constant N. Then it diverges by the comparison test.

For 11F, that function should work. Show (by contradiction), that there must be an integer k with 0 <= k < n and f(k/n)f((k+1)/n) < 0.
Alternatively for 11F, consider r=0n1f(rn) \sum_{r=0}^{n-1} f(\frac{r}{n}) . Directly you can then see that f(r/n)=0 for each r or at least one is positive and at least one is negative.
Reply 3
Original post by IrrationalNumber
Alternatively for 11F, consider r=0n1f(rn) \sum_{r=0}^{n-1} f(\frac{r}{n}) . Directly you can then see that f(r/n)=0 for each r or at least one is positive and at least one is negative.
Yeah, that's what I meant.
Original post by DFranklin
Yeah, that's what I meant.

Sorry, to be honest, I couldn't see how to use your hint (it looks like a product).
Reply 5
Original post by DFranklin
For 9F (iii): You should be able to see the terms behave like 1n\dfrac{1}{n} for large n. So you're expecting it to diverge. Since you're expecting it to diverge, the way forwards is to show that for large n the numerator is > kn^3 for some constant k, and the denominator is < Kn^4 for some other constant N. Then it diverges by the comparison test.

For 11F, that function should work. Show (by contradiction), that there must be an integer k with 0 <= k < n and f(k/n)f((k+1)/n) < 0.



Original post by IrrationalNumber
Alternatively for 11F, consider r=0n1f(rn) \sum_{r=0}^{n-1} f(\frac{r}{n}) . Directly you can then see that f(r/n)=0 for each r or at least one is positive and at least one is negative.


Thanks that worked.
Reply 6
Original post by IrrationalNumber
Sorry, to be honest, I couldn't see how to use your hint (it looks like a product).
f(a)f(b)<0 is just a "lazy" way of writing "one of f(a), f(b) is negative and one positive". I should have considered the equality case though.
Original post by IrrationalNumber
Alternatively for 11F, consider r=0n1f(rn) \sum_{r=0}^{n-1} f(\frac{r}{n}) . Directly you can then see that f(r/n)=0 for each r or at least one is positive and at least one is negative.



Very cute solution.
(edited 12 years ago)
Original post by DFranklin
f(a)f(b)<0 is just a "lazy" way of writing "one of f(a), f(b) is negative and one positive". I should have considered the equality case though.


For the first part of the question is one supposed to prove the Intermediate Value Theorem? Question 12 starts with "Prove Rolle's Theorem" which is clear but I am not sure what to do here.
Reply 9
Original post by worriedParnt
For the first part of the question is one supposed to prove the Intermediate Value Theorem? Question 12 starts with "Prove Rolle's Theorem" which is clear but I am not sure what to do here.


The first part of Q11 is essentially saying "prove the intermediate value theorem", except it's the special case where the intermediate value is 0 (which simplifies notation). So if you really want to then you could prove the full IVT, but it's not necessary.
Original post by nuodai
The first part of Q11 is essentially saying "prove the intermediate value theorem", except it's the special case where the intermediate value is 0 (which simplifies notation). So if you really want to then you could prove the full IVT, but it's not necessary.


Thanks very much

Quick Reply

Latest