Math Review Solutions to Homework #5 1. a) f(x) = 2 if x 0, and = x2 if x > 0. Note that f(0) = 2 and f(0 + ) = 0 for > 0 but “infinitely small”. It follows that, for any r > 0, there always exists an x Br(0) such that ||f(x) – f(0)|| = 2. Thus, there does not exist an r > 0 such that x Br(0) implies that ||f(x) – f(0)|| < e for all e > 0. This implies that f is not continuous at x = 0. Thus, f is not continuous on R. Not being continuous at x = 0, f is not differentiable at x = 0. Thus, f C1, and f C2. b) f(x) = 2 x if x 0, and = x2 if x > 0. Clearly, f is continuous and differentiable on R\{0}. Note that f(0) = 0 and f(0 + ) = 0 for “infinitely small”. It follows that there always exist an r > 0 such that x Br(0) implies that ||f(x) – f(0)|| < e for all e > 0. Thus, f is continuous at x = 0. We have Df(x) = 2, if x < 0 = 2 x, if x > 0. However, limx0 {[f(x) – 0 – A (x – 0)]/x} = limx0 {[f(x) – A x]/x} = limx0 [f(x)/x A] = 2 – A if x < 0 = – A if x > 0. Since we cannot have both (2-A) = 0 and A = 0, it follows that f is not differentiable at x = 0. Thus, f C1, and f C2. c) f(x) = 2 x2 if x 0, and = x2 if x > 0. Clearly, f is continuous and differentiable on R\{0}. Note that f(0) = 0 and f(0 + ) = 0 for “infinitely small”. It follows that there always exist an r > 0 such that x Br(0) implies that ||f(x) – f(0)|| < e for all e > 0. Thus, f is continuous at x = 0. We have Df(x) = 4 x, if x < 0 = 2 x, if x > 0. In addition, limx0 {[f(x) – 0 – A (x – 0)]/x} = limx0 {[f(x) – A x]/x} = limx0 [f(x)/x A] = – A if x < 0 = – A if x > 0. It follows that f is differentiable at x = 0, with A = Df(0) = 0 as derivative. Thus, f is differentiable on R. Clearly, Df is continuous and differentiable on R\{0}. Note that Df(0) = 0 and Df(0 + ) = 0 for “infinitely small”. It follows that there always exist an r > 0 such that x Br(0) implies that ||Df(x) – Df(0)|| < e for all e > 0. This implies that Df is continuous at x = 0, and therefore continuous on R. Thus, f C1. We have D2f(x) = 4, if x < 0 = 2, if x > 0. In addition, limx0 {[Df(x) – 0 – A (x – 0)]/x} = limx0 {[Df(x) – A x]/x} = limx0 [Df(x)/x A] = 4 – A if x < 0 = 2 – A if x > 0. Since we cannot have both (4 – A) = 0 and (2 – A) = 0, it follows that Df is not differentiable at x = 0. Thus, Df is not differentiable on R, and f C2. d) f(x) = 2 x3 if x 0, and = x4 if x > 0. Clearly, f is continuous and differentiable on R\{0}. Note that f(0) = 0 and f(0 + ) = 0 for “infinitely small”. It follows that there always exist an r > 0 such that x Br(0) implies that ||f(x) – f(0)|| < e for all e > 0. Thus, f is continuous at x = 0. We have 2 = 6 x2, if x < 0 = 4 x3, if x > 0. In addition, limx0 {[f(x) – 0 – A (x – 0)]/x} = limx0 {[f(x) – A x]/x} = limx0 [f(x)/x A] = – A if x < 0 = – A if x > 0. It follows that f is differentiable at x = 0, with A = Df(0) = 0 as derivative. Thus, f is differentiable on R. Clearly, Df is continuous and differentiable on R\{0}. Note that Df(0) = 0 and Df(0 + ) = 0 for “infinitely small”. It follows that there always exist an r > 0 such that x Br(0) implies that ||Df(x) – Df(0)|| < e for all e > 0. This implies that Df is continuous at x = 0, and therefore continuous on R. Thus, f C1. We have D2f(x) = 12 x, if x < 0 = 12 x2, if x > 0. In addition, limx0 {[Df(x) – 0 – A (x – 0)]/x} = limx0 {[Df(x) – A x]/x} = limx0 [Df(x)/x A] = – A if x < 0 = – A if x > 0. It follows that Df is differentiable at x = 0, with D2f(0) = A = 0. Thus, Df is differentiable on R. Clearly, D2f is continuous and differentiable on R\{0}. Note that D2f(0) = 0 and Df2(0 + ) = 0 for “infinitely small”. It follows that there always exist an r > 0 such that x Br(0) implies that ||f(x) – f(0)|| < e for all e > 0. This implies that D2f is continuous at x = 0. Thus, D2f is continuous on R, and f C2. Df(x) 2. (a) For x0 = 0, limxxo (x2 sin(1/x2) = 0 = f(x0) (since x2 0 and -1 sin(1/x2) 1). For x0 0, limxxo(x2 sin(1/x2)) = limxxo(x2) limxxo sin(1/x2) = x02 sin(1/x02) = f(x0). So f is continuous on R. (b) For x 0, f is continuously differentiable, since Df(x) = 2x sin(1/x2) - (2x3/x4) cos(1/x2) = 2x sin(1/x2) - (2/x) cos(1/x2) is continuous at x 0. Still, we need to show that f is differentiable at x = 0. From the definition of the derivative, at x=0, Df(x) = limx0 [(f(x)-f(0))/(x-0)] = limx0 [x2 sin(1/x2)/x] = limx0 (x sin(1/x2)) = 0, since |sin(1/x2)| 1. So, Df(0) = 0 exists. (c) limx0 Df(x) = limx0[2x sin(1/x2) - (2/x) cos(1/x2)] = limx0[2x sin(1/x2)] - limx0[(2/x) cos(1/x2). This limit does not exist at x = 0, since -1 cos(1/x2) 1, but (2/x) as x0. Thus, Df(x) is not continuous at x = 0. 3. (a) f (x, y) = x y ( x2 + y2 )-1/2 x2 y , (x, y) (0, 0) 2 ( x2 + y2 ) x2 + y f f(x,0) - f(0,0) 0-0 (0,0) = lim [ ] = lim [ ] = 0 , (x, y) = (0, 0) x 0 x 0 x x x - 3 f (x, y) = y x ( x2 + y2 )-1/2 xy2 , (x, y) (0, 0) 2 ( x2 + y2 ) x2 + y f(0, y) - f(0,0) f 0-0 (0,0) = lim [ ] = lim [ ] = 0 (x, y) = (0, 0). y0 y 0 y y y So the partial derivatives of f exist everywhere. (b) Notice that: f(x, y) - f(0, y) y f (0, y) = lim [ ] = lim [ ] = 1 or -1, for x = 0, y 0, 2 x 0 x0 x x x2 + y and f(x, y) - f(x,0) f x (x,0) = lim [ ] = lim [ ] = 1 or -1, for y = 0, x 0. 2 y0 y 0 y y x2 + y But: f f (0,0) lim y0 (0, y) = 1 x x f f 0 = (0,0) lim x 0 (x,0) = 1 y y 0= So Df is not continuous at (x, y) = (0, 0). (c) Assume that f is differentiable at (x, y) = (0, 0). f f Df(0,0) = ( , )(0,0) = (0, 0). x y x ||f(x, y) - f(0,0) - Df(0,0) || y = 0 lim (x,y) (0,0) ||(x, y) - (0,0)|| Choose x = y = a. This implies a ||f(a, a) - f(0,0) - Df(0,0) || a = 0, lim a0 ||(a, a)|| or a2 / 2 a2 ] = lim (1 / 2) = 0, lim [ a0 a0 2 a2 so we have a contradiction. Thus, f is not differentiable at (x, y) = (0, 0). 4. f(x) = a + b x + c x2 ||a + by + cy2 - (a + bx + cx2)|| ||f(y) - f(x)|| f (x) = lim [ ] = lim [ ] y x y x ||y - x|| ||y - x|| 4 ||b(y - x) + c(y2 - x2)|| = lim [ ] y x ||y - x|| = lim [b + c ( y x 5. y2 - x2 )] = lim (b + c (y + x)) = b + 2 c x . y x y-x 0 2 1 3/2 (a) D2f(x) = 0 4 x 2 2 0 1 D2f(x0) = at x0 = (1, 1) 0 4 The eigenvalues are 2 and -1/4, so the matrix is indefinite. 1 -3/2 1/2 1 -1/2 -1/2 x1 x 2 x 1 x 2 4 (b) D2f(x) = 4 1 1 -3/2 x 1-1/2 x -1/2 x 1/2 2 1 x2 4 4 1 1 4 at x0 = (1, 1). D2f(x0) = 4 1 1 4 4 The eigenvalues are 0 and -1/2, so the matrix is negative semi-definite. 2 x 22 4 x1 x2 (c) D2f(x) = 2 x 12 4 x 1 x 2 2 4 D2f(x0) = at x0 = (1, 1). 4 2 The eigenvalues are -2 and 6, so the matrix is indefinite. 1 -3/2 - x1 0 (d) D2f(x) = 4 1 0 - x -3/2 2 4 1 0 D2f(x0) = 4 at x0 = (1, 1). 1 0 - 4 The eigenvalues are (-1/4, -1/4), so the matrix is negative definite. 0 1 (e) D2f(x) = for all x. 1 0 The eigenvalues are (1, -1), so the matrix is indefinite. 5 0 0 0 (f) D2f(x) = 0 0 0 for all x. 0 0 0 The eigenvalues are (0, 0, 0), so the matrix is both positive semi-definite and negative semidefinite.