MA3110 Mathematical Analysis II AY 2010/2011 Sem 1 NATIONAL UNIVERSITY OF SINGAPORE MATHEMATICS SOCIETY PAST YEAR PAPER SOLUTIONS with credits to Agus Leonardi, Tay Jun Jie MA3110 Mathematical Analysis II AY 2010/2011 Sem 1 Question 1 (a) We will first prove supm≥n (am + bm ) ≤ supm≥n am + supm≥n bn . Let supm≥n am = L and supm≥n bm = M . By definition, for all m ≥ n, we have am ≤ L and bm ≤ M , which implies am + bm ≤ L + M for all m ≥ n, hence proving the claim. By the property of limits, we have lim sup (am + bm ) ≤ lim sup am + lim sup bn n→∞ m≥n n→∞ m≥n n→∞ m≥n hence proving the required inequality. (b) Let an = sin n and bn = − sin n. Then lim supn→∞ (an + bn ) = 0, while lim supn→∞ an = lim supn→∞ bn = 1. In this case, we have strict inequality. Question 2 (a) We will first prove that g 0 (0) exists implies g 0 (0) = 0. Suppose g 0 (0) exists, but g 0 (0) 6= 0. WLOG, suppose g 0 (0) > 0. Then there exists δ > 0 such that g(x) > 0, g(x) < 0, ∀x ∈ (0, δ) ∀x ∈ (−δ, 0) However, by definition, g(x) ≡ |f (x)| ≥ 0 for all x. Hence, we have a contradiction. We conclude that g 0 (0) = 0. Now, by definition of derivative, g(x) − g(0) g(x) |f (x)| = lim = lim =0 x→0 x→0 x x→0 x−0 x g 0 (0) = lim Moreover, when x > 0, we have the following inequality: − |f (x)| f (x) |f (x)| ≤ ≤ x x x The direction of above inequality is reversed when x < 0. Hence, by Squeeze Lemma, f (x) f 0 (0) = lim =0 x→0 x proving the statement. NUS Math LaTeXify Proj Team Page: 1 of 5 NUS Mathematics Society MA3110 Mathematical Analysis II AY 2010/2011 Sem 1 (b) Suppose f 0 (0) exists. One direction has been proven, i.e. if g 0 (0) exists then by part (i), f 0 (0) = 0 and g 0 (0) = 0. We will prove the other direction. Now, if f 0 (0) = 0, then by definition of derivative, given any ε > 0, there exists δ > 0, such that f (x) < ε. for all x such that 0 < |x| < δ, x Moreover, we have for all x such that 0 < |x| < δ, |f (x)| f (x) x − 0 = x < ε hence, proving that g 0 (0) = lim x→0 g(x) |f (x)| = lim =0 x→0 x x by definition. Question 3 (a) We will first show that f is Riemann-integrable in [0, 1], by showing that f is continuous in [0, 1]. Given ε > 0, take δ = ε2 , then for all |x − y| < δ, p √ |f (x) − f (y)| ≤ |y − x| < δ = ε as required. Now, we can apply the Riemann sum expression for f (x). Take the equal-spaced partition of f (with norm n1 ), with the right end-point of each partition interval as the sample point. Then since f is integrable, and limn→∞ n1 = 0, by the Theorem on Convergence of Riemann Sums, Z 1 n X i 1 f = f (x) dx n→∞ n n 0 lim i=1 as required. Question 4 (a) By Taylor Expansion, we have sin y = sin 0 + cos c y 1! for some c between 0 and y. xp xp As | cos c| ≤ 1, we have | sin y| ≤ |y| for all y ∈ R. Therefore, we have sin p ≤ p for all x ∈ R. n n To prove convergence of the required series, we will use Weierstrass M-test. Note that if p > 1, ∞ p ∞ X X x 1 = |xp | np np n=1 n=1 converges for all x ∈ R by the p-series test. Hence, ∞ X n=1 sin xp converges uniformly on [−r, r] for any r > 0 by Weierstrass M-test. np NUS Math LaTeXify Proj Team Page: 2 of 5 NUS Mathematics Society MA3110 Mathematical Analysis II AY 2010/2011 Sem 1 (b) By Taylor Expansion, we have sin y = sin 0 + cos 0 sin c 2 y− y 1! 2! for some c between 0 and y. 1 As −1 ≤ sin c ≤ 1, we have sin y ≥ y − y 2 for all y ∈ R. 2 Therefore, we have xp xp 1 x2p sin p ≥ p − n n 2 n2p Now, we will show that the series X ∞ ∞ p X 1 x2p 1 xp x 1 p − = − x np 2 n2p np 2 n2p n=1 n=1 diverges for any x 6= 0. X ∞ ∞ X 1 np − k k = − , where k = 12 xp . np n2p n2p n=1 n=1 np − k 1 Let an = and bn = p . n2p n Consider the series If p ∈ (0, 1], we have lim n→∞ an k = lim 1 − p = 1 > 0 n→∞ bn n Clearly, if p = 0, the series diverges. ∞ X Moreover, if p ∈ [0, 1], the series bn diverges by p-series test. n=1 Hence, by Limit Comparison Test, if p ∈ [0, 1], the series ∞ X sin n=1 xp is divergent at any x 6= 0. np Question 5 1 x sin √ . n n Clearly, fn (x) are differentiable for all x ∈ R. ∞ X Moreover, fn (1) converges. This is because (by Question 4) (a) Let fn (x) = n=1 1 1 sin √1 ≤ √ n n n n ∞ X 1 and the series 32 converges by the p-series test. n n=1 ∞ X Therefore, fn (1) converges by Comparison Test. n=1 Furthermore, ∞ X n=1 fn0 = ∞ X 1 x cos √ n n=1 n 3 2 converges uniformly. This is because 1 x 3 cos √ ≤ 13 2 n n2 n NUS Math LaTeXify Proj Team Page: 3 of 5 NUS Mathematics Society MA3110 and Mathematical Analysis II ∞ X 1 converges by p-series test. 3 n=1 AY 2010/2011 Sem 1 n2 Hence, the convergence of ∞ X fn0 follows by Weierstrass M-test. n=1 ∞ X 0 In this case, we have f (x) is uniformly convergent on R and f (x) = fn0 = n=1 ∞ X 1 x cos √ n n=1 n 3 2 as required. Question 6 √ ∞ ∞ X cos nx X 1 √ √ diverges. Furthermore, since cosine is even, it suffices to consider (a) For x = 0, = n n n=1 n=1 √ x > 0. Let x > 0 be given. Now, we note that cos nx ≥ 12 whenever π π √ ≤ nx ≤ 2kπ + , 3 3 2kπ − that is, (6k + 1)π 2 (6k − 1)π 2 ≤n≤ . 3x 3x 2 2 2 (6k+1)π Let lk = (6k−1)π . Since `k → ∞ as , L = , Ik = [lk , Lk ] and `k = Lk − lk = 8kπ k 3x 3x 3x2 k → ∞, we deduce that ∃N1 ∈ N such that Ik ∩ N 6= ∅ for every k ≥ N1 . In addition, let b·c, d·e and #Ik denote the floor function, ceiling function and number of integers in Ik respectively. Then #Ik = bLk c − dlk e + 1 > Lk − lk − 1 = `k − 1. Now, for every k ≥ N1 , √ √ X cos nx X cos nx 1 X 1 1 X 1 ≥ √ √ √ ≥ √ ≥ 2 2 n n n Lk n∈I n∈I n∈I n∈I k k k k 1 #Ik 1 `k − 1 = √ > √ 2 Lk 2 Lk Now, limk→∞ √1 Lk = 0 and 8kπ 2 `k 3x 4π = lim lim √ · = . k→∞ 3x2 k→∞ (6k + 1)π 3x Lk `√k −1 4π 2π −1 = 4π . Thus ∃N ∈ N such that − Hence, limk→∞ `√k L 2 3x 3x < 3x whenever k ≥ N2 . Therefore, Lk k for every k ≥ max(N1 , N2 ), we have √ X π cos nx √ > . n 3x n∈Ik √ ∞ X cos nx √ In conclusion, fails the Cauchy Criterion and thus diverges. n n=1 (b) For x = 0, √ ∞ X sin nx n=1 n = 0. Furthermore, since sine is an odd function, we have − √ ∞ X sin nx n=1 NUS Math LaTeXify Proj Team Page: 4 of 5 n = √ ∞ X sin (− nx) n=1 n , hence it NUS Mathematics Society MA3110 Mathematical Analysis II AY 2010/2011 Sem 1 suffices to consider the case where x > 0. Let x > 0 be given and let yn = √ nx for each n ∈ N. Note that ∞ [ [yn , yn+1 ) forms a parti- n=1 2 tion on [x, ∞). Moreover, limn→∞ yn+1 − yn = 0 and x2 = yn+1 − yn2 for all n ∈ N. Therefore, √ ∞ X sin nx n=1 n = x2 sin yn sin yn = (yn+1 + yn ) 2 (yn+1 − yn ). 2 yn yn √ ∞ X sin nx Since limn→∞ yn+1 −yn = 0, heuristically yn+1 ≈ yn for large n. Hence, the tail sum of n=1 Z 2 sin y is the Riemann sum of dy over some appropriate interval I. y I √ Z ∞ ∞ X sin t sin nx Since dt converges, we conclude that converges. t n 0 n n=1 NUS Math LaTeXify Proj Team Page: 5 of 5 NUS Mathematics Society
© Copyright 2024