Chain Rules, Taylor's Theorem and Function Extrema

Chain Rules, Taylor's Theorem and Function Extrema

Chain rules, Taylor’s theorem and function extrema I. Chain rules • Basic version (for one parameter). Suppose that h(t) = f(x1(t), . , xd(t)), where x1, ..., xd are differentiable at t and f is differentiable at (x1(t), . , xd(t)), then d ∂f h0(t) = X (x (t), . , x (t))x0 (t). (1) ∂x 1 d j j=1 j If x1(t), ..., xd(t) are differentiable functions of t and f is differentiable on the range of (x1(t), . , xd(t)), then (1) is sometimes expressed as dh ∂f dx ∂f dx = 1 + ··· + d . dt ∂x1 dt ∂xd dt – When d = 2 and h(t) = f(x(t), y(t)), where x and y are differen- tiable at t and f is differentiable at (x(t), y(t)), (1) becomes 0 0 0 h (t) = fx(x(t), y(t))x (t) + fy(x(t), y(t))y (t). (2) – Proof of (2). Recall that f is differentiable at (x0, y0) means that f(x, y) = f(x0, y0) + fx(x0, y0)(x − x0) + fy(x0, y0)(y − y0) +1(x − x0) + 2(y − y0), (3) where lim(x,y)→(x0,y0) 1 = 0 and lim(x,y)→(x0,y0) 2 = 0. Apply (3) with (x, y) = (x(t + ∆), y(t + ∆)) and (x0, y0) = (x(t), y(t)) and (2) holds by taking the limits as ∆ → 0. • Other versions. – Suppose that z(u, v) = f(x, y), where x = x(u, v) and y = y(u, v). Suppose that xu, xv, yu and yv exist and f is differentiable. Then ∂z ∂f ∂x ∂f ∂y ∂z ∂f ∂x ∂f ∂y = + and = + . ∂u ∂x ∂u ∂y ∂u ∂v ∂x ∂v ∂y ∂v – Suppose that u = f(x, y, z), where x = x(s, t), y = y(s, t) and z = z(s, t). Suppose that the partial derivatives of x, y and z with respect to s and t exist and f is differentiable. Then ∂u ∂f ∂x ∂f ∂y ∂f ∂z = + + ∂s ∂x ∂s ∂y ∂s ∂z ∂s 1 and ∂u ∂f ∂x ∂f ∂y ∂f ∂z = + + . ∂t ∂x ∂t ∂y ∂t ∂z ∂t Example 1. Suppose that h(t) = x2 + xy, where x = t and y = t2. Find h0(1). Example 2. Suppose that u = xy + 2z, where x = st, y = s + t and z = t2. Find ∂u/∂s. Example 3. Supposet that z = u + f(uv), where f is differentiable. Show that u(∂z/∂u) − v(∂z/∂v) = u. II. Taylor’s theorem • The multivariate version of Taylor’s theorem follows from the uni- variate Taylor’s theorem and the chain rule. To obtain the Taylor expansion of f(x) at a, let g(t) = f(a + t(x − a)), then g(0) = f(a) and g(1) = f(x). Then the univariate version of Taylor’s theorem says that ! n g(k)(0)tk g(n+1)(c)tn+1 g(t) = X + , k! (n + 1)! k=1 where c is between t and 0. Put t = 1 and we have the multivariate version of Taylor’s theorem. • Example 4. Suppose that f, fx and fy are differentiable on B((x0, y0), δ) for some δ > 0, where B((x0, y0), δ) = {(x, y): k(x, y) − (x0, y0)k < δ} and (x, y) ∈ B((x0, y0), δ). Show that R (x, y) f(x, y) = f(x , y ) + f (x , y )(x − x ) + f (x , y )(y − y ) + 2 , 0 0 x 0 0 0 y 0 0 0 2 where 2 R2(x, y) = fxx(x∗, y∗)(x − x0) + 2fxy(x∗, y∗)(x − x0)(y − y0) 2 +fyy(x∗, y∗)(y − y0) (4) and the point (x∗, y∗) lies on the segment with endpoints (x0, y0) and (x, y). • Taylor’s theorem can be applied to control the error of a linear ap- proximation if all the second-order partial derivatives are bounded. 2 III. Extrema of bivariate functions • Absolute extrema. – f(x, y) has an absolute maximum at (x0, y0) means f(x, y) ≤ f(x0, y0) for every (x, y) in the domain of f. – f(x, y) has an absolute minimum at (x0, y0) means f(x, y) ≥ f(x0, y0) for every (x, y) in the domain of f. • Relative extrema. Suppose that f is defined on a set containing B((x0, y0), δ) = {(x, y): k(x, y) − (x0, y0)k < δ} for some δ > 0. – f(x, y) has an relative maximum at (x0, y0) means f(x, y) ≤ f(x0, y0) for every (x, y) in B((x0, y0), δ). – f(x, y) has an relative minimum at (x0, y0) means f(x, y) ≥ f(x0, y0) for every (x, y) in B((x0, y0), δ). domain of f. • Partial derivative criteria for relative extrema (Theorem 11.11). Sup- pose that f is defined on a set containing B((x0, y0), δ) = {(x, y): k(x, y) − (x0, y0)k < δ} for some δ > 0. If f has a relative maximum or minimum at (x0, y0) and fx and fy both exist at (x0, y0), then fx(x0, y0) = fy(x0, y0) = 0. • If fx(x0, y0) = fy(x0, y0) = 0 or fx(x0, y0) and fy(x0, y0) do not both exist, then we say there is a critical point at (x0, y0). • Remark. If f is differentiable at (x0, y0), then f(x, y) ≈ L(x, y), where L(x, y) = f(x0, y0) + fx(x0, y0)(x − x0) + fy(x0, y0)(y − y0). We cannot have L(x, y) ≥ 0 on B((x0, y0), δ) or L(x, y) ≤ 0 on B((x0, y0), δ) unless fx(x0, y0) = fy(x0, y0) = 0. • Second partials test (Theorem 11.12). Suppose that fx(x0, y0) = fy(x0, y0) = 0 and all the second-order partial derivatives of f are continuous on B((x0, y0), δ) = {(x, y): k(x, y) − (x0, y0)k < δ} for 2 some δ > 0. Let D = fxxfyy − fxy. Then we have the following results. (a) If D(x0, y0) > 0, then (i) a relative minimum occurs at (x0, y0) if fxx(x0, y0) > 0 or fyy(x0, y0) > 0; 3 (ii) a relative maximum occurs at (x0, y0) if fxx(x0, y0) < 0 or fyy(x0, y0) < 0. (b) If D(x0, y0) < 0, then a saddle point occurs at (x0, y0). • We say a saddle point occurs at (x0, y0) if there is a critical point at (x0, y0) but there is no relative extremum at (x0, y0). 2 • The function D = fxxfyy − fxy is called the discriminant of f. • Remark. Under the conditions in the second partial test, when (x, y) ≈ (x0, y0), the R2(x, y) in (4) is close to 2 U(x, y) = fxx(x0, y0)(x − x0) + 2fxy(x0, y0)(x − x0)(y − y0) 2 +fyy(x0, y0)(y − y0) and U(x, y) f(x, y) ≈ f(x , y ) + . 0 0 2 2 Let D0 = fxx(x0, y0)fyy(x0, y0) − fxy(x0, y0). Then we have the fol- lowing results. (a) If D0 > 0, then (i) U(x, y) > 0 for (x, y) 6= (x0, y0) if fxx(x0, y0) > 0 or fyy(x0, y0) > 0; (ii) U(x, y) < 0 for (x, y) 6= (x0, y0) if fxx(x0, y0) < 0 or fyy(x0, y0) < 0. (b) If D0 < 0, then U(x, y) > 0 for some (x, y) ≈ (x0, y0) and U(x, y) < 0 for some other (x, y) ≈ (x0, y0). Example 5. Suppose that f(x, y) = x2 − y2. Find all the point(s) where the relative extrema or saddle point(s) of f occur. Answer: no relative extrema; saddle point at (0, 0) Example 6. Supppose that f(x, y) = x2 + y2 + 2x. Find all the point(s) where the relative extrema or saddle point(s) of f occur. Answer: relative minimum at (−1, 0). • Extreme value theorem for a bivariate function (Theorem 11.13). Sup- pose that f is a bivariate function that is continuous on S, where S is closed and bounded. Then f attains both its maximum and minimum on S. 4 – S is bounded means that there exists a number M > 0 such that S ⊂ {(x, y): k(x, y)k ≤ M}. – S is closed means that if limn→∞ k(xn, yn) − (x0, y0)k = 0 and ∞ {(xn, yn)}n=1 is a sequence in S, then (x0, y0) is also in S. – Extreme value theorem implies that the maximum and minimum of f on a closed and bounded set S can occur only at critical points or the boundary of S. – A point (x0, y0) is in the boundary of S means that for every δ > 0, B((x0, y0), δ) contains both some points in S and some points not in S. Example 7. Find the maximum and minimum of f(x, y) = x2 + y on S = {(x, y) : 0 ≤ x ≤ 1 and 2 ≤ y ≤ 3}. 5.

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    5 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us