Homework 5
Table of Contents
Exercise 10.2
Prove that bounded, decreasing sequences converge.
Solution.
Let (sn) be a bounded, decreasing sequence. Then the set {sn∣n∈N} is non-empty and bounded below, so s=infS exists. We will show that sn→s.
Let ε>0. Then s+ε is not a lower bound for (sn), so there exists N∈N such that sN<s+ε. Since (sn) is decreasing, we get
sn≤sN<s+εfor all n≥N.
Moreover, sn−s≥0 by definition of s, so for all n≥N,
∣sn−s∣=sn−s<ε.
Exercise 11.7
Let (rn) be an enumeration of the set Q of all rational numbers. Show there exists a subsequence (rnk) such that limk→∞rnk=+∞.
Solution.
I will give two solutions--one that just uses a theorem, and one that's more explicit in the construction. Either is fine, of course, but I suggest you understand the explicit construction.
Solution 1 (by Theorem 11.2)
Q is unbounded (e.g., N⊆Q), so because (rn) is an enumeration of Q, it follows that (rn) itself is unbounded. Thus, by THeorem 11.2(ii), there exists a subsequence which diverges to +∞.
Solution 2 (by construction)
Since (rn) is unbounded, 1 is not an upper bound for (rn). Thus, there exists n1∈N such that rn1≥1.
We proceed by induction. Suppose we have chosen n1<n2<⋯<nk such that rnk≥k. Notice that
Sk:=Q∖{r1,r2,…,rnk}
is still an unbounded set since we only remove finitely many elements. Thus, in particular, k+1 is not an upper bound for Sk, so there exists rnk+1∈Sk such that rnk+1≥k+1. By definition of Sk, we immediately have nk+1>nk as well.
Thus, we have shown that there exists a subsequence (rnk) such that rnk≥k, which implies rnk→+∞.
Exercise 11.9
- Show the closed interval [a,b] is a closed set.
- Is there a sequence (sn) such that (0,1) is its set of subsequential limits?
Solution.
-
Let (sn) be a convergent sequence of elements in [a,b] and let s=limn→∞sn. By definition, a≤sn≤b, so by the lemma from discussion, a≤s≤b, which means s∈[a,b].
-
No. Suppose there were such a sequence (sn). Recall that
n→∞limsupsn=sup(0,1)=1.
But we also showed that limsupn→∞sn itself is a subsequential limit, which means 1∈(0,1), which is impossible.