More Suprema and Density Practice

1.3.4, 1.3.5, 1.3.7, 1.3.9, 1.3.10, 1.4.6, 1.4.7

1.3.4 (check)

Question

Let A1,A2,A3,... be a collection of nonempty sets, each of which is bounded above.
a. Find a formula for sup(A1∪A2) and extend to ⋃k=1nAk
b. Consider sup(⋃k=1∞Ak). Does the formula in (a) extend to the infinite case?

Proof
a.

sup(A1∪A2)=max(supA1,supA2)

Proof
Let β=max(supA1,supA2). Clearly the claim holds if A1=A2 so say A1≠A2. Then either A1⊂A2 or A1⊃A2, which are argued in similar ways.

Using the former then A1∪A2=A2 so then we just have to show supA2≥supA1. Notice if they equal then we are done since sup(A1∪A2)=supA2=supA1=max(…). Now consider the case they don't equal. Then supA2−supA1>0 so then choose ε>0 to be that positive number. Since supA2 then there is a a∈A2 where supA2−ε=supA2−(supA2−supA1)=supA1<a. But since supA2 then ∀a′∈A2(a′≤supA2), thus giving that supA1<a≤supA2⇒supA1<supA2 as we wanted.
☐

Using an inductive proof shows the following with ease:

sup(⋃k=1nAk))=maxk=1nsupAk

b. It doesn't. Consider the collection A={Ai⊆R|Ai=[i,i+1]∀i∈N}. If you say that α=sup(⋃k=1∞Ai) exists then then ∃Ai such that α∈Ai. But then that means that α≥supAj for any j∈N. But then that means that α≥j for all j by the way we constructed Aj. This is a contradiction since that breaks the Cantor's Approach#^5404f9.
☐

1.3.5

Question

Let A∈R be nonempty and bounded above, let c∈R. This time define the set cA={ca|a∈A}.
a. If c≥0 then show that sup(cA)=csup(A)
b. Postulate a similar type of statement for sup(cA) for the case that c<0.

Proof
a. Clearly sup(cA) exists since A is nonempty and bounded above. Start from cα=csup(A) and we'll show that it is sup(cA). For the c=0 case notice that then cA={0} so clearly sup(cA)=0=0⋅sup(A). Thus, take c≠0:

  1. (cα is an upper bound for cA): Let a′∈cA be arbitrary. Then ∃a∈A where a′=ca. Then a′c=a. Since α=sup(A) then a≤α so consequently a′≤cα as desired.
  2. (cα is the least upper bound for cA): Let α′ be another upper bound for cA. Assume that α′<cα. Then α′c<α so then since α=supA then α′c∈A. Thus then ∃a∈A where α′c=a⇒α′=ca which contradicts α′ being an upper bound for cA.

b.

For c<0, then sup(cA)=−inf(A).

☐

1.3.7

Question

Prove that if a is an upper bound for A, and if a is also an element of A, then it must be that a=sup(A).

Proof
Clearly the upper bound for A property is satisfied so we just have to show that a is the least upper bound for A. Let a′ be another upper bound for A. Assume for contradiction that a′<a. Then since a∈A that contradicts a′ being an upper bound for A, so then clearly a′≥a as required.
☐

1.3.9 (check)

Question

a. If supA<supB, then ∃b∈B that is an upper bound for A.
b. Give an example to show that this is not always the case if we only assume supA≤supB.

Proof
a.
Since supA<supB⇒supB−supA=ε>0 so then by the Suprema Lemma for sup(B) then ∃b∈B such that supB−(supB−supA)=supA<b. Thus b is clearly an upper bound for A.

b. If A={q∈I|q2<2} and B={q∈Q|q2<2} then clearly supA=supB=2 but if there was some b∈B that's an upper bound for A then we know that b is irrational and 2 is irrational, so since The Density of Q in R#^ca78a0, then ∃q∈I where b<q<2, so clearly q∈A so clearly b is not an upper bound for A.
☐

1.3.10 (FINISH, don't forget the other parts)

(Cut Property)

If A,B are nonempty, disjoint sets with A∪B=R and a<b for all a∈A,b∈B then ∃c∈R such that x≤c whenever x∈A and x≥c whenever x∈B.

Proof

☐

1.4.7 (TODO)