Math 201-B Introduction to Proofs Instructor: Alex Roitershtein Iowa State University Department of Mathematics Fall 2015 Exam #3 (solutions to the practice test) December 18, 2015 Student name: - Student ID: Duration of the exam 120 minutes. The exam includes 5 questions. The total mark is 100 points. Please show all the work, not only the answers. Calculators, textbooks, and help sheets are allowed. 1. [20 points] (a) Let A and B be bounded subsets of R such that A ⊂ B. Show formally that sup A ≤ sup B and inf B ≤ inf A. (b) Let A = {x ∈ R : x = n−s for some n ∈ N, n ≥ 2 and s ∈ R, s > 0}. Show that sup A = 1 and inf A = 0. Solution: (a) Let a = sup A and b = sup B. By definition of b, b ≥ x ∀ x ∈ B. Since A ⊂ B, we obtain that b ≥ x for all x ∈ A. Thus b is an upper bound for A and hence b ≥ a = sup A. Similalrly, let c = inf A and d = inf B. By definition of d, d ≤ x ∀ x ∈ B. Since A ⊂ B, we obtain that d ≤ x for all x ∈ A. Thus d is a lower bound for A and hence d ≤ c = inf A. (b) For all n ∈ N, n ≥ 2 and s ∈ R, s > 0 n−s ≤ 2−s . 1 Consider now a function f (s) = 2−s , s > 0. Then log f (s) = −s log 2 ≤ 0. Thus sup A ≤ e0 = 1. On the other hand, if it would be the case that a := sup A ∈ (0, 1) we couldn’t find s > 0 such that −s log 2 > log a, log a or equivalently s < −log . Since log a < 0 and s > 0 is arbitrary, the last assertion is 2 clearly incorrect. This establishes that a = 1. On the other hand, while n−s > 0 for the parameters n, s with the range under the consideration, one can make n−1 (here we choose s = 1 for certainty) arbitrarily small by picking n ∈ N large enough. Thus (apply the result of (a) for included sets) 0 ≤ inf A ≤ inf{x ∈ R : x = n−1 for some n ∈ N, n ≥ 2} = 0. Thus sup A = 0. 2. [20 points] (a) Exercise 2.1.6 from the textbook BA. (b) Exercise 2.1.13 from the textbook BA. (c) Exercise 2.1.15 from the textbook BA. Solution: (a) We have: 0≤ 1 1 n ≤ ≤ . n2 + 1 n + n−1 n Since limn→∞ n−1 = 0, we obtain from the monotonicity of the limit that limn→∞ n n2 +1 = 0. (b) Since xn is monotone, say monotone increasing 9decreasing sequences are treated in a similar way), lim xn ≥ xm n→∞ for any m ∈ N. Therefore, if xm > xk for some m > k, then xk cannot be the limit of the sequence. This shows that xm = xk for all m > k. (c) The sequence is not bounded because the subsequence x2n+1 = 2n + 1 is not bounded. An example of a converging subsequence is x2n . Any its further subsequence is also converging. 3. [20 points] 2 (a) Exercise 2.2.3 from the textbook BA. (b) Exercise 2.2.8 from the textbook BA. Solution: (a) Use induction and the algebra of the limits. For k = 1 the result is trivia. On the otehr hand, if it is true for all integer j ≤ k, then lim xk+1 = n lim xkn · lim xn n→∞ k k+1 lim xn · lim xn = lim xn , n→∞ n→∞ = n→∞ n→∞ n→∞ as desired. (b) You are familiar with this kind of problems from Calculus II. Use, for instance, the L’Hospital’s Rule twice: n2 = n→∞ 2n x2 2x = lim x x x→∞ 2 x→∞ 2 log 2 2 = lim x = 0. x→∞ 2 (log 2)2 lim lim 4. [20 points] Exercise 2.1.8 from the textbook BA. Please argue using directly the definition of the limit. Solution: Let xn = 2n n! and Sn = 2n n=0 n! . Pn Then lim Sn = lim Sn−1 = n→∞ n→∞ ∞ X 2n n=0 n! = e2 . Hence lim xn = n→∞ lim (Sn − Sn−1 ) = lim Sn − lim Sn−1 n→∞ 2 n→∞ n→∞ = e − e2 = 0. 5. [20 points] Exercise 2.2.6 from the textbook BA. Solution: limn→∞ zn = 0 while limn→∞ wn = +∞ (thus the latter doesn’t exist in reals). Proposition 2.2.5 (iv) cannot be applied since limn→∞ xn = limn→∞ yn = 0 which rules out the set of conditions of the proposition. 3