Both sides previous revision
Previous revision
Next revision
|
Previous revision
|
math104-f21:hw4 [2021/09/16 10:45] pzhou created |
math104-f21:hw4 [2022/01/11 08:36] (current) pzhou ↷ Page moved from math104:hw4 to math104-f21:hw4 |
| |
5. Let (an) be a bounded sequence in R, and A=limsup(an), show that for any ϵ>0, the set {n∈N∣A−ϵ≤an≤A+ϵ} is infinite. | 5. Let (an) be a bounded sequence in R, and A=limsup(an), show that for any ϵ>0, the set {n∈N∣A−ϵ≤an≤A+ϵ} is infinite. |
| |
| ====== Solution ====== |
| 1. Suffice to show that, for any $\epsilon>0$, there is an N, such that for all n>N, we have |
| |a_n - x| \leq \epsilon, |
| which is |
| x−ϵ≤an≤x+ϵ. |
| By Cauchy condition for an, take ϵ1=ϵ, there exists N1>0, such that for all n,m>N1, we have ∣an−am∣<ϵ1, in particular, we have |
| am−ϵ≤an≤am+ϵ. |
| Consider the sequence (am)m=N1∞, then we have its formal limit satisfies, for all n>N1, |
| x−ϵ=(LIMam)−ϵ≤an≤(LIMam)+ϵ=x+ϵ. |
| Finally, take N=N1. |
| |
| 2. Take any N>2∣a∣, then for any n>N, we have |
| ∣an/an−1∣=∣a∣/n<1/2 |
| Hence $|a_{n}| \leq |a_{N}| (1/2)^{n-N}$, take limit n→∞ shows the result. |
| |
| 3. Since A is bounded above, we know the x=sup(A) exists. Suppose x2<3, then by the same argument as in Prop 5.5.12, we have a small enough $\epsilon>0$, such that (x+ϵ)2<3, then let q be a rational number, such that x<q<x+ϵ, we have q2<3, hence q∈A. This contradicts with with x being the sup of A. Suppose x2>3, then we can find small enough $\epsilon>0$, such that (x−ϵ)2>3, in particular x−ϵ is also an upperbound of A, hence contradict with x being the upper bound, hence x2=3. |
| |
| |
| 4. Since sn+12=1+sn, and (sn) converges, hence |
| limsn+12=lim(1+sn) |
| which gives |
| c2=1+c |
| solving this gives $c = (\pm \sqrt{5}+1)/2 . Since sn are all positive by induction and, we see $c = ( \sqrt{5}+1)/2$ |
| |
| 5. This is covered in last Friday's lecture, where we proved that L=limsup(an) is a limit point of an. Here we give the proof again. |
| |
| We prove by contradiction, suppose there is an $\epsilon>0$, such that there are only finitely many terms in (an) contained in $[L-\epsilon, L+\epsilon].ThenthereexistsanN_1,suchthatforalln>N_1,|a_n - L| > \epsilon.Bypropertyof\limsup,weknowtherealsoexistsN_2,suchthatforalln > N_2,a_n < L + \epsilon.Then,forn > N= \max(N_1, N_2)$, we have |
| |a_n - L| > \epsilon, \quad a_n < L + \epsilon \Rightarrow a_n < L - \epsilon. |
| Taking limsup on the left, we get L \leq L - \epsilon$, which is a contradiction. |
| |
| |
| |
| |
| |
| |