Introduction Directional Derivative of a Scalar

Total Page:16

File Type:pdf, Size:1020Kb

Introduction Directional Derivative of a Scalar Classroom Tips and Techniques: Directional Derivatives in Maple Robert J. Lopez Emeritus Professor of Mathematics and Maple Fellow Maplesoft Introduction The directional derivative of a scalar function , computed in Cartesian coordinates, is defined by , where is the vector x evaluated along a line with direction u. When it exists, it can be evaluated as . In Maple, it is easy to implement both the definition and the evaluation, as we will show. Several identities in vector calculus involve the operator acting on a vector B. The resulting expression is interpreted as the directional derivative of the vector B in the direction of the vector A. This is not easy to implement in Maple's VectorCalculus package. However, this functionality exists in the Physics:-Vectors package, and in the DifferentialGeometry package where it is properly called the DirectionalCovariantDerivative. This month's article explores all of these issues. Directional Derivative of a Scalar Cartesian Coordinates Top Level If is a unit vector in Cartesian coordinates, then the rate of change of the scalar function taken at in the direction U is = as computed by Maple. This directional derivative can be written as , or as , provided we define as the vector . VectorCalculus The DirectionalDiff command in the VectorCalculus packages will compute, at , the directional derivative of the scalar function in the direction of the vector V, where V does not have to be a unit vector. We illustrate these calculations for the Student VectorCalculus package because this package is a bit more lenient with respect to defining coordinate systems and the names of coordinate variables. Tools≻Load Package: Student Vector Calculus Loading Student:-VectorCalculus Define the vector . Apply the DirectionalDiff command. Note that Maple normalizes V. The DirectionalDiff command supports evaluation at a point from within the command itself. Obtain the directional derivative from first principles Apply ∇ to obtain the gradient of f. Apply the Normalize command to V. Normalize V by dividing it by its length. Student MultivariateCalculus The Student MultivariateCalculus package has a DirectionalDerivative command that normalizes the direction vector and also has the "evaluation" property. Tools≻Load Package: Student Multivariate Calculus Loading Student:-MultivariateCalculus Apply the DirectionalDerivative command. Physics:-Vectors The Physics:-Vectors package also has a DirectionalDiff command that will normalize the direction vector. However, it does not have the "evaluation" property. Restart Maple. Load the Physics:-Vectors package. Allow notational enhancements. Define the direction vector V. The basis vector i can be entered as or as . Apply the DirectionalDiff command. The vector V is normalized by DirectionalDiff. Polar Coordinates In Cartesian coordinates, the defining calculation = suggests that when the limit of this derivative exists, it can be more easily calculated as . The paradigm extends to other coordinate systems. VectorCalculus Restart Maple. Tools≻Load Package: Student Vector Loading Student:-VectorCalculus Calculus Define the direction vector V as a vector field in polar coordinates. Apply the DirectionalDiff command to The third argument indicates the system in which to interpret . Simplify: , and Change to , etc. Maple obtains this result by mapping the whole calculation back to Cartesian coordinates, applying the paradigm , then mapping back to polar coordinates. Note that Maple also normalizes the vector V. Physics:-Vectors Like the Student VectorCalculus package, the Physics:-Vectors package can work in Cartesian and , as well as spherical and cylindrical coordinates. For polar coordinates, Physics:-Vectors uses the cylindrical system with basis vectors and . The VectorCalculus package itself works with all of the orthogonal coordinate systems known to Maple. (See for the complete list.) Restart Maple. Load the Physics:-Vectors package. Allow notational enhancements. Define the direction vector V in polar coordinates. The basis vector can be entered as or as . Apply the DirectionalDiff command. The vector V is normalized by DirectionalDiff. This result is equivalent to the result produced by the DirectionalDiff command in the VectorCalculus packages. Directional Derivative of a Vector The operator acting on a vector B appears in the vector identities for the gradient of a dot product, and for the curl of a cross product. In particular, the operator appears in the identities and and is generally interpreted as the directional derivative of one vector taken in the direction of the other vector. Indeed, it is a form of the directional covariant derivative of differential geometry. Just as the directional derivative of a scalar leads to the definition of the gradient vector, so too does the directional derivative of a vector lead to the covariant derivative, a rank-two tensor. Implementing this directional derivative of a vector in the confines of the VectorCalculus packages pushes these packages to their limits. In effect, the calculation can only be done in Cartesian coordinates, and that, by an artifact. Outside its DifferentialGeometry package, Maple cannot form the operator , so must first compute the gradients of each component of B, and then dot A with each such gradient. The Physics:-Vectors package extends the applicability of its DirectionalDiff command, but unfortunately normalizes the direction vector when applied to a scalar (correct) and also when applied to a vector (problematic). Cartesian Coordinates VectorCalculus Restart Maple. Tools≻Load Package: Student Vector Calculus Loading Student:-VectorCalculus Set display format for vectors. Define vector fields F and V. Compute by mapping the gradient onto the components of F, then mapping the dot product with V onto each component of that "vector." Compute by mapping DirectionalDiff onto the components of F. This then requires multiplying by because DirectionalDiff normalizes the direction vector V. The syntax for mapping the DirectionalDiff command onto the components of F is far more complex than that required for mapping both the gradient and the dot product. This is for two reasons. First, the DirectionalDiff command normalized the direction vector (here, V); and second, because the Norm command in the VectorCalculus packages produces a function whose value at a point is the norm of the given vector. A proposal to modify DirectionalDiff in the VectorCalculus packages has already been made. Under this proposal, application of the command to a vector field would be allowed, and when so applied, the direction vector would not be normalized. Maplesoft has given serious consideration to this modification, and if implemented, would allow the (2-D) Math notation to stand for . Physics:-Vectors Although the DirectionalDiff command in the Physics:-Vectors package automatically maps onto the components of a vector, the direction vector is normalized. Unlike the VectorCalculus packages, the Physics:-Vectors package lets the syntax to form a valid operator that can be applied to a vector B. Hence, the notation will compute the directional derivative of B in the direction of A. Restart Maple. Load the Physics:-Vectors package. Allow notational enhancements. Define the vector F. The basis vector i can be entered as or as . Define the direction vector V. The basis vector i can be entered as or as . Apply the DirectionalDiff command, organizing the result by components. Implement the notation , again organizing the result by components. Since the DirectionalDiff command again normalizes the direction vector, the best implementation of the directional derivative of a vector is via the notation in the Physics:-Vectors package, at least in Cartesian coordinates. Polar Coordinates As earlier, polar coordinates are used as an example of nonCartesian coordinates. VectorCalculus In the VectorCalculus packages, the mapping artifact that works in Cartesian coordinates does not work in polar coordinates. A wrong answer results. Therefore, to obtain the directional derivative of a vector in polar coordinates, the calculation must be implemented in Cartesian coordinates. Restart Maple. Tools≻Load Package: Student Vector Loading Student:-VectorCalculus Calculus Set display format for vectors. Define, in polar coordinates, the vector fields F and V. Apply the MapToBasis command to convert F and V to vector fields in Cartesian coordinates. Obtain the directional derivative. Map the result back to polar coordinates. Simplify: , and Change to , etc. Display the directional derivative, after having grouped the result by components. In this directional derivative, the terms and in the first and second components respectively, imply that cannot be represented in nonCartesian coordinates as an operator acting componentwise on F. This is because ultimately, derivatives have to be taken of the varying basis vectors and . Indeed, let , , define a line through the Cartesian point in the direction . The polar vector F evaluated along this line is The directional derivative is given by , in which the basis vectors and vary along the line described by . Thus, computing requires differentiating the products and , a process that would generate the Christoffel symbols of covariant differentiation. It is this differentiation process that is responsible for the non-differentiated terms in the directional derivative, and therefore the need to carry out the calculation in Cartesian coordinates.
Recommended publications
  • The Directional Derivative the Derivative of a Real Valued Function (Scalar field) with Respect to a Vector
    Math 253 The Directional Derivative The derivative of a real valued function (scalar field) with respect to a vector. f(x + h) f(x) What is the vector space analog to the usual derivative in one variable? f 0(x) = lim − ? h 0 h ! x -2 0 2 5 Suppose f is a real valued function (a mapping f : Rn R). ! (e.g. f(x; y) = x2 y2). − 0 Unlike the case with plane figures, functions grow in a variety of z ways at each point on a surface (or n{dimensional structure). We'll be interested in examining the way (f) changes as we move -5 from a point X~ to a nearby point in a particular direction. -2 0 y 2 Figure 1: f(x; y) = x2 y2 − The Directional Derivative x -2 0 2 5 We'll be interested in examining the way (f) changes as we move ~ in a particular direction 0 from a point X to a nearby point . z -5 -2 0 y 2 Figure 2: f(x; y) = x2 y2 − y If we give the direction by using a second vector, say ~u, then for →u X→ + h any real number h, the vector X~ + h~u represents a change in X→ position from X~ along a line through X~ parallel to ~u. →u →u h x Figure 3: Change in position along line parallel to ~u The Directional Derivative x -2 0 2 5 We'll be interested in examining the way (f) changes as we move ~ in a particular direction 0 from a point X to a nearby point .
    [Show full text]
  • Notes for Math 136: Review of Calculus
    Notes for Math 136: Review of Calculus Gyu Eun Lee These notes were written for my personal use for teaching purposes for the course Math 136 running in the Spring of 2016 at UCLA. They are also intended for use as a general calculus reference for this course. In these notes we will briefly review the results of the calculus of several variables most frequently used in partial differential equations. The selection of topics is inspired by the list of topics given by Strauss at the end of section 1.1, but we will not cover everything. Certain topics are covered in the Appendix to the textbook, and we will not deal with them here. The results of single-variable calculus are assumed to be familiar. Practicality, not rigor, is the aim. This document will be updated throughout the quarter as we encounter material involving additional topics. We will focus on functions of two real variables, u = u(x;t). Most definitions and results listed will have generalizations to higher numbers of variables, and we hope the generalizations will be reasonably straightforward to the reader if they become necessary. Occasionally (notably for the chain rule in its many forms) it will be convenient to work with functions of an arbitrary number of variables. We will be rather loose about the issues of differentiability and integrability: unless otherwise stated, we will assume all derivatives and integrals exist, and if we require it we will assume derivatives are continuous up to the required order. (This is also generally the assumption in Strauss.) 1 Differential calculus of several variables 1.1 Partial derivatives Given a scalar function u = u(x;t) of two variables, we may hold one of the variables constant and regard it as a function of one variable only: vx(t) = u(x;t) = wt(x): Then the partial derivatives of u with respect of x and with respect to t are defined as the familiar derivatives from single variable calculus: ¶u dv ¶u dw = x ; = t : ¶t dt ¶x dx c 2016, Gyu Eun Lee.
    [Show full text]
  • Tensor Calculus and Differential Geometry
    Course Notes Tensor Calculus and Differential Geometry 2WAH0 Luc Florack March 10, 2021 Cover illustration: papyrus fragment from Euclid’s Elements of Geometry, Book II [8]. Contents Preface iii Notation 1 1 Prerequisites from Linear Algebra 3 2 Tensor Calculus 7 2.1 Vector Spaces and Bases . .7 2.2 Dual Vector Spaces and Dual Bases . .8 2.3 The Kronecker Tensor . 10 2.4 Inner Products . 11 2.5 Reciprocal Bases . 14 2.6 Bases, Dual Bases, Reciprocal Bases: Mutual Relations . 16 2.7 Examples of Vectors and Covectors . 17 2.8 Tensors . 18 2.8.1 Tensors in all Generality . 18 2.8.2 Tensors Subject to Symmetries . 22 2.8.3 Symmetry and Antisymmetry Preserving Product Operators . 24 2.8.4 Vector Spaces with an Oriented Volume . 31 2.8.5 Tensors on an Inner Product Space . 34 2.8.6 Tensor Transformations . 36 2.8.6.1 “Absolute Tensors” . 37 CONTENTS i 2.8.6.2 “Relative Tensors” . 38 2.8.6.3 “Pseudo Tensors” . 41 2.8.7 Contractions . 43 2.9 The Hodge Star Operator . 43 3 Differential Geometry 47 3.1 Euclidean Space: Cartesian and Curvilinear Coordinates . 47 3.2 Differentiable Manifolds . 48 3.3 Tangent Vectors . 49 3.4 Tangent and Cotangent Bundle . 50 3.5 Exterior Derivative . 51 3.6 Affine Connection . 52 3.7 Lie Derivative . 55 3.8 Torsion . 55 3.9 Levi-Civita Connection . 56 3.10 Geodesics . 57 3.11 Curvature . 58 3.12 Push-Forward and Pull-Back . 59 3.13 Examples . 60 3.13.1 Polar Coordinates in the Euclidean Plane .
    [Show full text]
  • Matrix Calculus
    Appendix D Matrix Calculus From too much study, and from extreme passion, cometh madnesse. Isaac Newton [205, §5] − D.1 Gradient, Directional derivative, Taylor series D.1.1 Gradients Gradient of a differentiable real function f(x) : RK R with respect to its vector argument is defined uniquely in terms of partial derivatives→ ∂f(x) ∂x1 ∂f(x) , ∂x2 RK f(x) . (2053) ∇ . ∈ . ∂f(x) ∂xK while the second-order gradient of the twice differentiable real function with respect to its vector argument is traditionally called the Hessian; 2 2 2 ∂ f(x) ∂ f(x) ∂ f(x) 2 ∂x1 ∂x1∂x2 ··· ∂x1∂xK 2 2 2 ∂ f(x) ∂ f(x) ∂ f(x) 2 2 K f(x) , ∂x2∂x1 ∂x2 ··· ∂x2∂xK S (2054) ∇ . ∈ . .. 2 2 2 ∂ f(x) ∂ f(x) ∂ f(x) 2 ∂xK ∂x1 ∂xK ∂x2 ∂x ··· K interpreted ∂f(x) ∂f(x) 2 ∂ ∂ 2 ∂ f(x) ∂x1 ∂x2 ∂ f(x) = = = (2055) ∂x1∂x2 ³∂x2 ´ ³∂x1 ´ ∂x2∂x1 Dattorro, Convex Optimization Euclidean Distance Geometry, Mεβoo, 2005, v2020.02.29. 599 600 APPENDIX D. MATRIX CALCULUS The gradient of vector-valued function v(x) : R RN on real domain is a row vector → v(x) , ∂v1(x) ∂v2(x) ∂vN (x) RN (2056) ∇ ∂x ∂x ··· ∂x ∈ h i while the second-order gradient is 2 2 2 2 , ∂ v1(x) ∂ v2(x) ∂ vN (x) RN v(x) 2 2 2 (2057) ∇ ∂x ∂x ··· ∂x ∈ h i Gradient of vector-valued function h(x) : RK RN on vector domain is → ∂h1(x) ∂h2(x) ∂hN (x) ∂x1 ∂x1 ··· ∂x1 ∂h1(x) ∂h2(x) ∂hN (x) h(x) , ∂x2 ∂x2 ··· ∂x2 ∇ .
    [Show full text]
  • Math 1320-9 Notes of 3/30/20 11.6 Directional Derivatives and Gradients
    Math 1320-9 Notes of 3/30/20 11.6 Directional Derivatives and Gradients • Recall our definitions: f(x0 + h; y0) − f(x0; y0) fx(x0; y0) = lim h−!0 h f(x0; y0 + h) − f(x0; y0) and fy(x0; y0) = lim h−!0 h • You can think of these definitions in several ways, e.g., − You keep all but one variable fixed, and differentiate with respect to the one special variable. − You restrict the function to a line whose direction vector is one of the standard basis vectors, and differentiate that restriction with respect to the corresponding variable. • In the case of fx we move in the direction of < 1; 0 >, and in the case of fy, we move in the direction for < 0; 1 >. Math 1320-9 Notes of 3/30/20 page 1 How about doing the same thing in the direction of some other unit vector u =< a; b >; say. • We define: The directional derivative of f at (x0; y0) in the direc- tion of the (unit) vector u is f(x0 + ha; y0 + hb) − f(x0; y0) Duf(x0; y0) = lim h−!0 h (if this limit exists). • Thus we restrict the function f to the line (x0; y0) + tu, think of it as a function g(t) = f(x0 + ta; y0 + tb); and compute g0(t). • But, by the chain rule, d f(x + ta; y + tb) = f (x ; y )a + f (x ; y )b dt 0 0 x 0 0 y 0 0 =< fx(x0; y0); fy(x0; y0) > · < a; b > • Thus we can compute the directional derivatives by the formula Duf(x0; y0) =< fx(x0; y0); fy(x0; y0) > · < a; b > : • Of course, the partial derivatives @=@x and @=@y are just directional derivatives in the directions i =< 0; 1 > and j =< 0; 1 >, respectively.
    [Show full text]
  • Derivatives Along Vectors and Directional Derivatives
    DERIVATIVES ALONG VECTORS AND DIRECTIONAL DERIVATIVES Math 225 Derivatives Along Vectors Suppose that f is a function of two variables, that is, f : R2 → R, or, if we are thinking without coordinates, f : E2 → R. The function f could be the distance to some point or curve, the altitude function for some landscape, or temperature (assumed to be static, i.e., not changing with time). Let P ∈ E2, and assume some disk centered at p is contained in the domain of f, that is, assume that P is an interior point of the domain of f.This allows us to move in any direction from P , at least a little, and stay in the domain of f. We want to ask how fast f(X) changes as X moves away from P , and to express it as some kind of derivative. The answer clearly depends on which direction you go. If f is not constant near P ,then f(X) increases in some directions, and decreases in others. If we move along a level curve of f,thenf(X) doesn’t change at all (that’s what a level curve means—a curve on which the function is constant). The answer also depends on how fast you go. Suppose that f measures temperature, and that some particle is moving along a path through P . The particle experiences a change of temperature. This happens not because the temperature is a function of time, but rather because of the particle’s motion. Now suppose that a second particle moves along the same path in the same direction, but faster.
    [Show full text]
  • Directional Derivatives
    Directional Derivatives The Question Suppose that you leave the point (a, b) moving with velocity ~v = v , v . Suppose further h 1 2i that the temperature at (x, y) is f(x, y). Then what rate of change of temperature do you feel? The Answers Let’s set the beginning of time, t = 0, to the time at which you leave (a, b). Then at time t you are at (a + v1t, b + v2t) and feel the temperature f(a + v1t, b + v2t). So the change in temperature between time 0 and time t is f(a + v1t, b + v2t) f(a, b), the average rate of change of temperature, −f(a+v1t,b+v2t) f(a,b) per unit time, between time 0 and time t is t − and the instantaneous rate of f(a+v1t,b+v2t) f(a,b) change of temperature per unit time as you leave (a, b) is limt 0 − . We apply → t the approximation f(a + ∆x, b + ∆y) f(a, b) fx(a, b) ∆x + fy(a, b) ∆y − ≈ with ∆x = v t and ∆y = v t. In the limit as t 0, the approximation becomes exact and we have 1 2 → f(a+v1t,b+v2t) f(a,b) fx(a,b) v1t+fy (a,b) v2t lim − = lim t 0 t t 0 t → → = fx(a, b) v1 + fy(a, b) v2 = fx(a, b),fy(a, b) v , v h i·h 1 2i The vector fx(a, b),fy(a, b) is denoted ~ f(a, b) and is called “the gradient of the function h i ∇ f at the point (a, b)”.
    [Show full text]
  • Lecture Notes of Thomas' Calculus Section 11.5 Directional Derivatives
    Lecture Notes of Thomas’ Calculus Section 11.5 Directional Derivatives, Gradient Vectors and Tangent Planes Instructor: Dr. ZHANG Zhengru Lecture Notes of Section 11.5 Page 1 OUTLINE • Directional Derivatives in the Plane • Interpretation of the Directional Derivative • Calculation • Properties of Directional Derivatives • Gradients and Tangents to Level Curves • Algebra Rules for Gradients • Increments and Distances • Functions of Three Variables • Tangent Panes and Normal Lines • Planes Tangent to a Surface z = f(x, y) • Other Applications Why introduce the directional derivatives? • Let’s start from the derivative of single variable functions. Consider y = f(x), its ′ ′ derivative f (x0) implies the rate of change of f. f (x0) > 0 means f increasing as x ′ becomes large. f (x0) < 0 means f decreasing as x becomes large. • Consider two-variable function f(x, y). The partial derivative fx(x0,y0) implies the rate of change in the direction of~i while keep y as a constant y0. fx(x0,y0) > 0 implies increasing in x direction. Another interpretation, the vertical plane passes through (x0,y0) and parallel to x-axis intersects the surface f(x, y) in the curve C. fx(x0,y0) means the slope of the curve C. Similarly, The partial derivative fy(x0,y0) implies the rate of change in the direction of ~j while keep x as a constant x0. • It is natural to consider the rate of change of f(x, y) in any direction ~u instead of ~i and ~j. Therefore, directional derivatives should be introduced. ... Lecture Notes of Section 11.5 Page 2 Definition 1 The derivative of f at P0(x0,y0) in the direction of the unit vector ~u = u1~i + u2~j is the number df f(x0 + su1,y0 + su2) − f(x0,y0) = lim , ds s→0 s ~u,P0 provided the limit exists.
    [Show full text]
  • Multivariable Calculus Review
    Outline Multi-Variable Calculus Point-Set Topology Compactness The Weierstrass Extreme Value Theorem Operator and Matrix Norms Mean Value Theorem Multivariable Calculus Review Multivariable Calculus Review Outline Multi-Variable Calculus Point-Set Topology Compactness The Weierstrass Extreme Value Theorem Operator and Matrix Norms Mean Value Theorem Multi-Variable Calculus Point-Set Topology Compactness The Weierstrass Extreme Value Theorem Operator and Matrix Norms Mean Value Theorem Multivariable Calculus Review n I ν(x) ≥ 0 8 x 2 R with equality iff x = 0. n I ν(αx) = jαjν(x) 8 x 2 R α 2 R n I ν(x + y) ≤ ν(x) + ν(y) 8 x; y 2 R We usually denote ν(x) by kxk. Norms are convex functions. lp norms 1 Pn p p kxkp := ( i=1 jxi j ) ; 1 ≤ p < 1 kxk1 = maxi=1;:::;n jxi j Outline Multi-Variable Calculus Point-Set Topology Compactness The Weierstrass Extreme Value Theorem Operator and Matrix Norms Mean Value Theorem Multi-Variable Calculus Norms: n n A function ν : R ! R is a vector norm on R if Multivariable Calculus Review n I ν(αx) = jαjν(x) 8 x 2 R α 2 R n I ν(x + y) ≤ ν(x) + ν(y) 8 x; y 2 R We usually denote ν(x) by kxk. Norms are convex functions. lp norms 1 Pn p p kxkp := ( i=1 jxi j ) ; 1 ≤ p < 1 kxk1 = maxi=1;:::;n jxi j Outline Multi-Variable Calculus Point-Set Topology Compactness The Weierstrass Extreme Value Theorem Operator and Matrix Norms Mean Value Theorem Multi-Variable Calculus Norms: n n A function ν : R ! R is a vector norm on R if n I ν(x) ≥ 0 8 x 2 R with equality iff x = 0.
    [Show full text]
  • Derivatives in N-Dimensional Spaces
    5 DERIVATIVES IN N-DIMENSIONAL VECTOR SPACES A Introduction , -' ,.- * -..: The approach of our textbook (as d&ioped in the previous two units) in emphasizing the geometrical aspects of partial derivatives, directional derivatives, and the gradient is, in many ways, an excellent approach. We get a picture of what appears to be happening, and, especially for the beginning student, this is much more meaningful than a more precise and general discussion that would apply to all vector spaces, regardless of dimension. In fact, if we want to use the latter approach, it is wise to use the former first. This is in accord with many such decisions which have already been made .in our previous treatment of mathematics. As an example, within the framework of our present course, notice how carefully we ex- ploited the geometric aspects of vectors as arrows before we even mentjoned n-dimensional vector spaces. And, in this respect, recall how sophisticated many of our "arrow" results were in their own right. That is,.granted that arrows were simpler than n-tuples, when all we had were arrows, there were still some rather difficult ideas for us to assimilate. In any event, our aim in this chapter is to continue within the spirit of the game of mathematics which, thus far, has served us so well. Namely, we have introduced n-dimensional vector spaces so that we could utilize the similarity in form between the previously studied functions of the form f(x) and our new functions of the form f(x). For example, when it came time to define continuity and limits for functions of several real variables, we rewrote f(xl, ...,xn) as f (5) and then mimicked the f (x) situation by copying definitions verbatim, subject only to making the appropriate "vectorizations." Rather than review this procedure, let us instead move to the next plateau.
    [Show full text]
  • Generalized Directional Derivatives and Subgradients of Nonconvex
    ( )i,\ , . t, ., \\\rt to.o j -i .1.. GENERT\LIZED DIRECTIO\AL DERI\TATIVES AND SUBGRADIENTS OF NONCONVEX FUNCTIONS R T. ROCKAFIJLLAI{ l lntroduction. Srudjes of optimjzation prollens and ccrlah kiD(ls of difierentiaL eqLrations hale led in rcccnt )enrs to thc dclelopnrent of a gen cralized rheor] of dift-erentiaiion qLrite distincr iD spirit ancl rarse of applicati,m from the onc base.l on L. Schwari:z's 'distriblriions. ' This thco( associares .l-r,,t-\. u" t, r. .. o ., ;r . oto,og: .r, .. ,. L point r a I certain elements of the clLral space E+ caLLed \tLy tints tit gtarlitnt js stneralircd s ot .f at '. 1 lcsc form A set dI(r) rhat a]lrat.s conre: and Ncak*-closed L'possiLl_v cmprl ). The nrlrifurctior df : r + bflx) is thc Rules that relate Dl to scncraljzed djrectional cterilatives of /, or:rlloN d,i 1o b€ expr.ssed or estimarcd ;n terns of thc sLrbcljlTerenrjaLs of other tunctjoos (\i hen.l : ,h + f'. f = ! o ,1, etc ), co prise the vrrd.r'fl./.rrill ar/.&/,s. Such rrLes arc used espcciall! ir anall,zing ihe con.Liijon 0 a D.f(s), Nhich i]picalt,v Deans that:f is some sort of'quasi-optlnaL'poinr ior,l The cxtcnded-real- laluccl nrtlrc of,i is essential in such a contcxr as a delicc for representins SulijlTerenrial caiclils bcgan rrith com-cx lLrrctions on R". Rockafeilar l43l defincd d,f ior su.h funcriorrs, shon'cd hor t.} characicrjze A_ft]j) nr rerms of one sidcd cLirectional rleriratiles _l'l:;r r), and prolecl that rndcr mild restrictions rLlles sllcI .s (1.1) d(.n +/,)(r) : d,no) + dj,ir) are valj.l.
    [Show full text]
  • Derivation of the Formula for a Directional Derivative
    Derivation of the Directional Derivative Formula We let ~u = u1~{ + u2~| be a fixed unit vector in the xy-plane, and as usual, we let z = f(x; y). We want to show that the simple formula for the directional derivative f~u = fxu1 + fyu2 is valid wherever f is a differentiable function. We fix a particular point (a; b) in the xy-plane where f is differentiable, and we recall that the formal definition of fu(a; b) is f(a + u1h; b + u2h) f(a; b) fu(a; b) = lim − ; h 0 h ! so we want to show that the limit simplifies to fx(a; b)u1 + fy(a; b)u2 : ( ) ∗ Now we can certainly rewrite the difference quotient in the definition in the following form, dividing it into one part where the x-value varies while y remains fixed and the other where the reverse holds: f(a + u1h; b + u2h) f(a; b) f(a + u1h; b + u2h) f(a; b + u2h) f(a; b + u2h) f(a; b) − = − u1 + − u2 : ( ) h u1h u2h ∗∗ Because the change in the x-value in the first term is u1h, we have arranged for the denominator to have that value, in hopes that it will approach the partial derivative of f with respect to x; and similarly for y in the second term. Now as h 0, we also have u2h 0, so the fraction [f(a; b + u2h) f(a; b)]=(u2h) ! ! − in the second term goes to fy(a; b). And the factors u1 and u2 are constants.
    [Show full text]