CHAPTER 11 EE PARTIAL DIFFERENTIATION Tre book of nawure is written in the language of mathematics." PREVIEW ‘This chapter extends the methods of single-variable differential calculus to functions of (Wo oF more independent variables. You will learn how 0 take derivatives of such functions and to interpret those derivatives as slopes and rates of change. You will also, learn a more general version of the chain rule and extended procedures for optimizing 1 function, The vector methods developed in Chapters 9 and 10 play an important role in this chapter, and indeed, we will find that the closest analogue of the single-variable Uerivative is a certain vector function called the gradient. In physics, the gradient is the rte at which a variable quantity, such as temperature or pressure, changes in value. In this chapter, we will define the gradient of a function, which is a veetor whose components along the axes are related (0 the rate at which the function changes in the direction of the given component. PERSPECTIVE In many practical situations, the value of one quamity depends on the values of «wo or more others. For example, the amount of water in a reservoir depends on the amount tof rainfall and on the amount of water consumed by local residents and thus may be regarded as a function of two independent variables. The current in an electrical circuit js function of four variables: the electromotive force, the capacitance, the resistance, and the inductance. We will analyze a variety of models using techniques and tools that will be developed in this chapter. is quan Is ually acute 0 Galion, bt sco (othe Hsia Rebeka Higgit wos probably wston by Chet Chi ne of Calin end, sappy aa fests, CONTENTS 14.4 Functions of Sever Level euves and surfaces Graphs of funtion of 90 ies 44.2 Limits and Continuity ‘Open and closed ses i and Limit of function of wo varahles Continuity Limi an contig for fonctions of tive variables 11 Partial Derivatives aril differentiation Parial derivative a a slope Paria derivative a arate Higher-order partial derivatives 144 Tangent Planes, Approximatio Ditterontiability “Tangent planes Incremental upproximations The tou diferesuil Ditferecabiity 44.8 Chain Rulon ‘Chain rule for one parameter Exionsions of the ehain rule 14.8 Directionsl Derivatives and the Gradient The ditectiona derivative The pra Moximal property of the " Punetions of three vacihl nd Normal property of tbe sraient, “Tangent planes and normal lies 147 Extrema of Functions of Two Varlablee Relative ext Second partials test, Absolute extrema of continous functions Least squotes approximation af data 14.8 Lagrange Multiplors ‘Method of grange uh Constrained aptimization robles Lagrange multiplies with 90 Parameters A geomecicalimerpretaion of lgrange’s theorem Chapter 11 Review Research Project, 819820 14.1. Functions of Several Variables 11.1 FUNCTIONS OF SEVERAL VARIABLES IN THIS SECTION: Basic concepts, level curves and surfaces, graphs of functions of two variables ‘The basic terminology of functions of several variables is introduced, along with the notion of level curves and surfaces. We also examine various methods for sketching graphs of functions of 1wo variables. Basic Concepts Physical quantities often depend on two or more variables (see the Perspective at the beginning of this chapter). For example, we might be concerned with the temperature, TT, at various points (x,y) on a metal plate, In this situation, T might be regarded as a function of the two location variables, « and y. Extending the notation for a function of wwe can denote this relationship by T(x,y). FUNCTION OF TWO VARIABLES A function of two variables iss rule / that assigns 10 cach ordered pair (4) in a set Da unique number f(x, ¥). The set D is called the domain of the function, and the corresponding values of f(x, ¥) constitute the range of f. Functions of three or more variables can be defined in a similar fashion, For example, the temperature in our introductory example may vary aot only on the plate, but also with lime 1, in which case it would be denoted by T(r,,),1). Occasionally, we will examine functions of four or more variables, but for simplicity, we will Focus most of our attention oon functions of 1wo or three variables. When dealing with a function f of wo variables, we may weite z = f(x,y) and reter to. and y as the independent variables and (o z as the dependent variable, Ofien, the functional “rule” will be given as formula, and unless otherwise stated, we will assume that the domain is the largest set of points in the plane (or in RB) for which the functional formua is defined an real-valued. These definitions and conventions are illustrated in Example | for a function of two variables. Example 1 Domain, range, and evaluating a funetion of two variables Let f(x,y) = VO HF 8, Evaluate 2,1) and f(t b, Describe the domain and range of f Solution a fQ1)=V9=F=4 One) 4(ey Wa =-woa 3 b. The domain of f is the set of all ordered pairs (x,y) for which YO—¥? — Hy? is 9, in order for 9 an? defined. We must have 9 = x? = 4y? > 0, or, equivalently, x? + 4y f the square root 10 be defined. Thus, the’ domain of fis the set of all points ¢r,) inside or on the ellipse x? +4)? = 9, as shown in Figure 11.1, The range of f isthe Figure 11.1 The domain for set ofall numbers z = 9 =x —4y? for (x,y) in the domain x? + 4y? <9. Thus, the given function the range is the interval 0 =z <3. Functions of several variables can be combined in much the same way as functions ofa single variable, as noted in the following definition14.4 Functions of Several Variables A polynomial funetion in « and y is « sum of functions of the form cy" ‘with nonnegative integers im and 2 and C a constant; for example, aay 78y 2x ay FU polynomial in x and y. A rational funetion is a cuoticnt of a polynomial function ivided by a nonzero polynomial function. Similar notation and terminology apply Sanetions of three or more variables, Level Curves and Surfaces By analogy withthe single-variable case, we define the graph of the funetion f(e,y) to be the collection of all 3-tuples—ordered triples—(x, y,z) such that (8,9) 8 im the domain of f and z =/(x,3). The graph of f(x,y) isa Surface in R? whose projection ‘onto the xy-plane is the domain D. Tis usually not easy to sketch the graph of @ function of wo variables without the assistance of technology." One way to proceed is illustrated in Figure 11.2. When we sraphed quadric surfaces in Section 9.7 we used the zrace of @ graph in a plane, and we vse that ides here, Love] curse of fat 8 y Level curve of fat 5 Figure 11.2 Graph of a function of two variables Notice that when the plane z = C intersects the surface < = F(r,y), the result is the ace, the equation f(r.) = C. The set of points (x,y) in the xy-plane that satisfy Sc he Compatona Window athe ow ofthis ton, 82114.1. Functions of Several Variables F(e.y) = C is called the level curve of f at C, and an entire family of level curves is generated as C varios over the range of f. We can think of a trace as a “slice” of the surface at a particular location anc! a level curve as its projection onto the xy-plane. By sketching members of this family on the xy-plane, we obtain a useful topographical map of the surface z =/(,y). Because level curves are used to show the shape of fa surface (a mountain, for example, as shown in Figure 11.3), they are sometimes called contour curves. , The surface as a mountain c. Level curves yield a topographic map 4. Topographic map of Mt, Rainier Figure 11.8 Level curves of a surface For example, imagine that the surface z = f(x, )) is a “mountain,” and that we wi to draw a two-dimensional “profile” of its shape. To draw such a profile, we indicate the paths of constant elevation by sketching the family of level curves in the plane and pinning to each curve ta show the elevation (© which it corresponds, as shown in11.1 Functions of Several Variables 823 Figure 11.3e, Notice that regions in the map where paths are crowded together correspond to the steeper portions of the mountain. An actual topographical map of Mount Rainier is shown in Figure 11.34. ‘You probably have seen level curves on the weather report in the newspaper or television news shows, where level curves of equal temperature are called isotherms (see Figure 11.4), Other common uses of level curves involve curves of equal pressure (called jisobars) and curves of electric potential (called equipotential lines). Figure 11.4 Isotherms Example 2. Level curves Sketch some level curves of the funetion f(x,y) = 10 — Solution computer graph of 2 =/¢x.y) is the surface shown in Figure 11.Sa, Figure 11,5b shows traces of the graph of fin the planes z = 1, z = 6, and z = 9, and the corresponding level curves are shown in Figure 11.Se. A table of values is shown in the margin, TR ARN TOS fa Interactive Computer generated graph. Simplified graph showing traces ¢. Level curves of surface Figure 11.5 Three ways of viewing the surface z = 10 —824 11.1 Functions of Several Variables Graphs of Functions of Two Variables ‘The level curves of a function f(x,y) provide information about the cross sections of the surface z = f(0,y) perpendicular to the z-axis. However, a more complete of the surface cam often be oblained by examining ctoss sections in other directions as ‘well. This procedure is used to graph a funetion in Example 3. Example 3. Level curves Use the level curves of the function (x,y) = 12 +52 to sketch the graph of Solution The level curve.x* + y? = 0 (that is, C = 0) 5s the point (0,0), and for C > 0, the level curve x? + y? = C is the circle with center (0,0) and radius VC (Figure 11.6a) ‘There are no points (x.y) that satisfy x? y= C for C <0. ceEE: Interactive The level curves are circles . The cross sections perpendicular to the ‘y-axis are parabolas (shown in white) Figure 11.6 The graph of the function f(x,y) = 4? + We can gain additional information about the appearance of the surface by examining cross sections perpendicular to the other two principal directions; that is, 10 the x-axis and the y-axis, Cross-sectional planes perpendicular to the x-axis have the form x =A and intersect the surface +9? in parabolas of the form z 3? For example, the cross section of the plane x = 5 intersects the surface in the parabola z = 25 +y° That is itis the set of points (5, y,25 + y2) as y-varies, Similarly, cross-sectional planes perpendicular to the y-axis have the form y = B and intersect the surface in parabolas (oF the form z = x24 8? (shown in white in Figure 11.6b). To summarize, the surface z = x? + y® has cross sections that are circles in planes perpendicular to the z-axis (Figure 16a) and parabolas in the other two principal diree- tions. Since the surface is formed by revolving a parabola about its axis, itis called & circular paraboloid or a paraboloid of revolution. The concept of level curve can be generalized to apply to functions of more than wo variables. In particular, if f is a function of three variables x, y, and <, then the solution set of the equation /(x,),2) = C is a region of R? called a level surface of f at C11.1 Functions of Several Variables 825 COMPUTATIONAL WINDOW Computer software is available for graphing functions of two variables, Software packages such as Mathematica, Maple, Matiab, and Derive will sketch very sophisticated graphs in three dimensions. Software usually allows a aiven graph to be visualized from different viewpoints by rotating the surface until a desired viewpoint is found. A few examples of computer generated surfaces are shown in Figure 11.8, 1 seer Figure 11.7 Interactive Computer generated surfaces, Seny) = Fey) = sin Example 4 Isothermal surface Suppose 2 region of B? by Tr y,2) = 100 = 04 T>0, is heated so that its temperature T at each point (x.y.2) is given y? ~ 2¥ degrees Celsius, Describe te isothermal surfaces for Solution The isothermal surfaces are given by T(x.y,<) =k far constant Ay that is x2 + 100 =k, If 100~& > 0, the graph of x? 4 )2-4 2? = 100—k is a sphere of radius VIOD=E and center (0,0,0). When k= 100, the graph is a single point (the origin), and 7(0,0,0) = 1OO°C. As the temperature drops, the constant & gets smaller, and the dius 700 ~ € ofthe sphere gets larger, Hence, the isothermal surfaces ave spheres, and the lager the radius, the cooler the surface. This situation is illustrated in Figure 11.7826 14.1. Functions of Several Variables re VW E r= VIG ‘Spores with larger radi correspond to eaoler temperatures, Figure 11.8 Isothermal surfaces for T As you can see by the examples in this section, you will need to recall the graphs of quadric surfaces (see Section 9.7). It will also help if you can recognize the surface by looking at its equation. What you will need to remember is summarized in Table 9.2, page 736. PROBLEM SET 11.1 Find the dematn and range foreach fein given in 1, What does this say? Discuss what is meant by e fone- — Pr@blens 5-14, tion of (wo variables. Your discussion should include ‘examples of a function of two variables not discussed 8. f(ry) = VEEP inthis section, 1 6 fy) = 2, MWhat does this say? Deserbe how level eurves ver cain be used to sketch the graph ofa function (3). flusu) = J Include some examples in your discussion (diferent = from hos ine ex, Bros fe 3. Letfs.y) = 32) ay? It ris areal numb, fn 9. for.y) =Inty =) 2 f0.0) BfH10) & fO=1) fad efOd RFD 10, fee) = Virsing ear) End 11. ft.) = VOIP FO DF 4, Letfte.y2) =a ye 4 ty = a. £0,040) b FL-1D) pony) a ehh) a Spee) 1B. fee) M4, for) a ad © Bit. d ihe11.1 Functions of Several Variables Sketch some level curves f(x,y) = C for © > Ofor each function given in Problems 15-20 1S. f(s.y 2x ~3y 16. fry) 17. f(nsy) = 18. gee 19. Atxy) =? + 20. ftx.y) = y In Problems 21-26, sketch and describe the level surface F 6,942) = C for the given value of C. 2 fle.y.2) 2? forC = 1 22. flxyys2) = 42 +2? for C 23. flx.ys2) ety — fore 24, fix.y.2) =¥y =z lore =0 25. flx,y.2) = + DP + —2P + ~ 3) fore =4 26. f(x,y.) = 2? = 2y? =z for = 1 In Problems 27-34, deseribe the iraces of the given quadric surface in each coordinate plane, then sketch the surface. 27, 9x? Ay? 827 ‘Match each family of level curves given in Problems 35-40 with one of the surfaces labeled A-E. 3S. foyyaxt—y? 36. f(x,y)c D. Skeich the graph of each fiction given in Problems 41-82, AL Foy) B. Soy) HB L697 41 FO) 9. Fo) 42 sory 44. F659) 46. forsy) = 4B soe 50. fy) = VERT SL flx.y) 174)? +2 82, for.y) = To? 53, IF E(x.) isthe voltage (potential) at each point (x,y) inthe plane, the level curves of E are called equipoten- tid curves. Suppose Fwy) Sketch the equipotential curves that correspond to Fale =2,andE 54, According tothe ideal gas law, PY = kT, where P is pressure, Vis volume, 7 is temperature, and k is eon stant, Suppose a tank contains 3500 in. of a gas ata pressure of 24 bin? when the temperatue is 270°K ‘ogeees Kelvin, 1, Determine the constant of proportionality k. be, Express the temperature 7 as a function of P and V using the constant found in part a and sketch the isotherms for the temperature Functions 88. The fens equation in optics states that 11.1. Functions of Several Variables where dis the distance of an olject from s thin, spher- ical lens, dis the distance of its image on the other side of the lens, and L is the focal lent of the lens (see Figure 11.9). Write Las a function od, and dy and describe some level curves of the function (these are curves of constant focal length), loge Figure 11.9 Image of an object through a lens 56, ‘The EZGRO sericultural company estimates that when 100 worker-hours of labor are employed on y acres of land, the number of bushels of wheat produced is JFoe,y) = Axty”, where A, a, and b are nonnegative ‘constants. Suppose the company decides to double the production factors x and y. Determine how this deci- sion affects the production of wheat in each of these matb>| baths! Gabba! 57. Suppose that when « machines and y worker-hours we used cach day, a certain fuctory will produce QCx,y) = [Oxy celiphones, Describe the relationship between the “inputs” x and y that result in an “output” ‘of 1,000 phones each day. Note: You are Finding a level ceurve of the production function Q. 88. A publishing house fas found that in a certain city each of its salespeople will sell approximately T0005 units per month, at a price of p dollars/unit, where s denotes the total number of salespeople employed and. ‘isthe amount of money spent each month on local advertising. Express the total revenue R as. function of poryand . 59, Modeling Problem A manufacturer with exclusive ‘rights 10 sophisticated new industrial machine is plan ning to sella limited number of the machines 10 both foreign and domestic firms. The price that the man- ulactorer can expect to receive for the machines will depend on the number of machines made av“41.2 Limits and Continuity to the domestic m rarer, the mac! ell for putis modeled by 829 ‘ket and y' machines to the foreign 60, Modeling Problem At a certain factory, the daily out- Q=cK'L units, where K denotes capital investment, Lis the size of the labor force, and C and r are constants, with <4 < (this is called a Cobb-Douglas production thousand dollars apiece abroad. Express the revenue R function ). What happens to Q if K and L are both dou- ‘sa function of x and y. bled? Wi 11.2 LIMITS AND CONTINUITY i both are wipled? IN THIS SECTION: Open and closed sets in ? and &, mit of a function of two variables, continuity, limits and continuity for functions of three variables As in the single-variable case, the concept of limit plays a fundamental role in the calculus of functions of sev. variables, Our focus is on functions of two variables, but the notions discussed also apply to functions of three or more variables, Most commonly considered funet described in terms of intervals, However, dealing with functions of two or more variables requires special terminology and notation that we will introduce in this section and then use to discuss limits and continuity for functions of two variables. Open and Closed Sets in P? and 2? An open disk centered at the poi that C(a,b) im B? is the set of all points P¢e,y) such Vea FO =) 0 (see Figure 11.108). If the boundary of she disk fs included ¢that is, it J@ a+ GBP S 1), the disk is said to be elosed (sce Figure 11.10b). Open and Closed disks are analogous to open and closed intervals on a coordinate line. A poim Po is said to be an interior point ofa set $ in R? if some open disk centered al Po is contained entirely within $ (as shown in Figure 11.11a). If § is the empty set, cor if every point of $ is an interior point, then 5 is called an open set (Figure 11.110). A point Po is called a boundary point of $ if every open disk centered at Pp contains botiv points that belong to and points that do not. The collection of all boundary points of § is called the boundary of S, and S is said to he closed if it contains its boundary (Figure 11.116). The empty set and R? are both open and closed, Similarly, an open ball centered at C(a,b,c) is the set of all points (x,y,z) such that = ay + —bP + =e 0. A point Py is an interior point of 1 set S in BR? if there exists an open ball centered at Py that is contained emtirely wi 5, and the nonempty set $ is an open set if all its points are interior points. A point Py is a boundary point of S if every open disk centered at My contains both points that belong to $ and points that do not, and S is closed if it contains all its boundary points ‘As with BA, the set 2? is both open and closed Oey +4. a, Open disk 44+ b. Closed disk Figure 11.10 Disks in the ayrplane830 11.2 Limits and Continuity I ‘Boundary ' \ i | (oi Bounty i hy ; point | \ a, Interior poims and boundary bb, The sot $ is open if alts points _¢. The set S is closed if it contains points of a sot are interior points all its boundary points Figure 11.11 Open and closed points in R Limit of a Function of Two Variables In Chapter 2, we informally defined the lint statement lim f(x) = L to mean that f¢«) ean be made arbitrarily close to F. by choosing x sufficiently close (but not equal) 10 e. For a function of two Variables we write (yak eo Baot ? to mean that the functional values f(c,y) ean be made arbitrarily close to the number L by choosing the point (x,y) sufficiently close to the point (40.0). The following is a ‘more formal definition of this limiting behavior. Wha sis sys Ur Hi G69) = La then given € > 0, the funtion value of f(x,y) must lie in the interval (L— ¢,L+€) whenever (sy) is @ point in the domain of f other than P(xa,so) that lies inside the disk of radius 8 centered at Po, This is illustrated in Figore 11.12. (Cake wih sais 4a with deleted emer (493 A punctured disk b, Sash tim tert got Figure 11.12. Limit of @ function of two variables44.2 Limits and Continuity 831 When considering inn f(x) we need (0 examine the approach of « to ¢ from so irections (the Jef and Tight-hand limits). However, for @ function of two variables, we write (x,y) + G.y0) to mean that the point (x,y) is allowed to approach (xp, 0) along ‘any curve in the domain of f that passes through (19, 9). For example, no matter how you get (0 the 15th step of Denver's state capitol, you will still be at 5,280 ft when you {got there, IT the limit im few (ag)! =L is not the same for all approaches, or paths within the domain of f, then the limit does ot exist. Example 1 Ev ating a limit of a function of two variables tim LRT Ayo Solution rst, note that for x # 3%. ct be-» y Therefore, lin o+b= Example 2 Showing init doesnot exist 2 Way) = along, colon’? s does not exist by evaluating this limit yee Solution First note that the denominator is zero at (0,0) s0 (0,0) is not defined. IF ‘we approach the origin along the x-axis (where y = 0), we find that ng the x-axis, the y-axis, and along the tine Tie 2x(0) : +0 } 10 i Figure 11.13 Graph of a= PY and limits a8, +e 50 fx») > 0 as (x,y) > (0,0) along y = 0 (and x # 0). We find similar result if ‘we approach the origin along the y-axis (where x =0); see Figure 11.13, Gay) + (0.0) However, along the line y = x, the functional values (for x 7 0) are so that > 1 as Gr.) > (0,0) along y =x, Because f(r.) tends toward dif- ferent numbers as (x,y) > (0,0) along different curves, it follows that F has no limit at the origin,is often possible ro show that Timi does not exist by she methods Hustrated in Esaples Zand 3 However, is inpossible ro sry 12 rove ta him, fy) exis boy showing the finding value of Fox.9) isthe sane along every ‘uve tar passes through tx9,80) Because there are too many such 14.2 Limits and Continuity lim, eno does not exist Solution If we approach this problem as we did the one in Example 2, we would take limits as (x,y) > (0,0) along the x-axis, the y-axis, and the line y =x. All these approach lines have the general form y = mx. We find: lim tm 32 coo tHO0) Ry? (so nO ATE Gua? ms = litte Frage ter too) YE me =0 Along y = ms: However, if we approach (0,0) slong the parabola y = x2, we find: ey tin, te = tim, crater) x (coder x? + (7) = tim bey Det ‘Therefore, since approaching (0,0) along the lines y = mx gives o different limiting value from approaching along the parabola y =.17, we conclude that the given limit does not exist 8 ‘We have just observed tha iti difficult to show that given limit exists, However, limits of functions of two variables that are known to exist ean be manipulated in much he same way as limits involving functions of a single variable, Here is a list of the basic rules for manipulating such bn14.2 Limits and Continuity 833 Example 4 Manipulating limits of functions of two variables Assuming each im Tim, 0? +49 +97) wots Solution ae tim, 6? kay £99) = PG) + OF aga =3 A graph is shown below Fini, 2ay geo 2) Tim Gey) 20) PHP 4 A graph is shown below. Continuity Recall that a function of # single variable x is continuous atx = ¢ if 1. (6) is defined, 2. fim £0) exisixs 3. fim 49 =U) Using the definition of the limit of a function of two variables, we ean now define the continuity of a function of two variables analogously.414.2 Limits and Continuity CONTINUITY AS A FUNCTION OF TWO VARIABLES The function f(x.) is continuous at the point (xg, sp) if and only if nn 1. (60.90) i defined: Jim £049) exists: retusa! Tim” £09) =FC0.90). Sas) ear ‘The function f is continuous on a set § if it is continuous at each point in S. st The function f is continuous at (xo, yo) if the functional value of f(x,3) is close to (2, yo) whenever (r,’) in the domain of fis sufficiently close a). Geometrically, this means that f is continuous if the surface z = f(x,y) “holes” or “gaps” (see Figure 11.14), 8. A “hole” discontinuity A “gap” discontinuity Figure 11.14 Two types of discontinuity, “The basic properties of limits can be used to show that iff and g are both continuous fon the set S, then so are the sum f + g, the multiple af, the product fg, the quotient /¢ (whenever g #0), and the root YF wherever it is defined. Also, if F is a function of ‘wo variables, that is F (io, 0) continuous at (i, jo) and G is a function of one variable ‘that is continuous at x9 (oF yo), it can be shown that the composite function Ge F is continuous at (89,30). Many common functions of two variables are continuous wherever they are defined, For instance, a polynomial in two variables, such as x3y?+ 3xy? — 7x +2, s continuous throughout the plane, and a rational function in two variables i continuous Wherever the denominator polynomial is not zero. Unless otherwise stated, the functions of two oF more variables considered in this text will be continuous wherever they are defined. Example $ Testing for continuity ‘Fest the continuity of the functions: xay a £09) fery ‘These are the functions whose graphs are shown in Figure 11.14.14.2 Limits and Continuity Solution a. The function f is a rational funetion of x and y (since x — y and x? +9? are both polynomials), so itis discontinuous only where the denominator is 0; namely at (0,0). 1b. Again, this is a rational function and is discontinuous only where the denominator is 0} that is, where y 0 ‘Thus, the function is continuous at all points « copt those lying on the parabol yer. Example 6 Continuity using the limit definition Show that f is continuous at (0,0) where a, Front . Side view Figure 11.18 Graph of f Solution The function appears to be continuous. Since the function vanishes along the y-axis, we only need to analyze it at points (x,y) with x #0. To prove continuity at (0,0), we must show that for any € > 0, there exists a 8 > 0 such that [F099 = £10,001 = [Fry] < € whenever 0 <7 #y? < x #0 (We use y? and 3* here instcad of y/** + y? and § for convenience.) [fx = 0, then Lfe.y) = 0. Ix #0 note that Lf (2,9)] = ly because [sin +] = 1 for & #0. Thus, in cither case, [x.y] < ly]. Also, if (x,y) lies in the disk x? y? < 6%, then the points (0,y) that satisly y? < 8° lie in the same disk (let.x 2 6), In other words, points satisfying y| <6 lic in the disk, and if we let 5 = ¢ it follows that U7. 9) ~F10.0)] 5 by] <8 = whenever 0 < fx? +» Limits and Continuity for Functions of Three Variables The concepts we have just intreduced for functions oF two variables in (o functions of three variables in B®. In particular, the Finit statement oxtend naturally lim fony Gaya) bonnetPROBLEM SET 11.2 11.2 Limits and Continuity means that for each number ¢ > 0, there exists a number 5 > 0 such that Ifeay.z) =H] <€ whenever (x,7,£) is a point in the domain of f such that 0< Vea TO Fea <8 The function fv, '.2) is continuous at the point Polo, Yo. 0) if 1. f (40.30.20) is defined; 2 Him yf Code) exists: 3, fim Fesy2) =F Goro. 20) Most commonly considered functions of three variables are continuous wherever they sre defined, We elose with an example that illustrates how to determine the set of dis- ‘continuities of a function of three variables. Example 7 Continuity for a funetion of three variables For what points (x,y,2) is the following function continuous? Solution “The function f(r.y.2) is continuous except where it is not defined; that is for x+y? — 22 <0. Thus, f(x,y.) is continuous at any point not inside or on the paraboloid ¢ = Hx? +y") ls or insofar of wo ee 9. tm, [em tim vy +5 10. tim Inge? +92 career y+5) swat OE PD 4 (5x? = 2ey +? +3) 2 (x? = 1)? - 4) vole, @= Dy 2)11.2 mits and Continuity 837 ef tan"! 25. me 26. lim (e+ ye 1. tim ‘ Le 10.0) wotn® + txoh Mao) . 2p 1 tim weston Fa 15, tim e429 estan a Fy seta ins + £089) 27. im, (sins — cosy) ais aconstant 1. 20, im ceo tom Joey? a2 28 Hn, | In Problems 21-30, evaluate the indicated limit, if it ene00) exists. fit does not exist, give reasons. fim, (sy? +239 tim | (Sx? — Dey by" Fe Hin FED A ih B28 +9) 23 lim tela ey 30, tim tat 00 [ cost? + kyIn Problems 31-34, show that exist BL feey) mf Gey) does not cea = fans “eae BB. F(0,9) for (x, ¥) #00) for (x.y) = @.0) Is f continuous at (0,0)? Explain. 36. Let f be the function defined by for (x.y) # (0,0) fawy=) Ee 0 for (a3 Is f continuous at (0, 0)? Explain. 37. Let f be the function defined by (0,0) x29? feet t for (v.99 # .0) Find “lim, Fer. oo) tay b. Prove that has no limit at (0,0). 38. Let f be the function defined by Foy) = For (x.y) 0.0) a Find 14.2 Limits and Continuity 39, Given that the function 3x3 = 399 fora? #97 a otherwise reoe| co is continuous atthe origin, what is A? 40, Given that the function for (x,y) # (0,0) for (x.y) = (0,0) is continuous atthe origin, what is B? 41. Given that the function (° is continuous atthe origin, what is C? Hint: Use polar coordinates, 2. Let xty? +2)? for (ry) # 0.0) (0,0) Fee) for (yy (a? 97) Sony AS for (x,y) # (0,0). For what value of f(0,0) is (6.3) continuous at 0.0)? Hint: Use polar coordinates, 43. Let for (xy) # (0,0) Sys 0 for (x,y) = (0,0) Given that fc.) has a limit at (0,0), is the function f continuous there’? 44, Let for (x.y) # (0,0) four) = | +y 0 for (x,y) = (0,0) Given that f(x.) has a limit at (0,0), is the Funetion f continuous there? 45. Assuming that the limit exists, show that y lim, (ya) 00m Use polar coordinates to find the limit given in Prob- ems 46-49. ‘ ceo! Mom 25 48, tim csr}11.3 Partial Derivatives 49. coho’ BYE Think Tank Problems tn Prablemns 80°52, eter show 50. Ti lim f(0,y) = O,then tim, for.y) cst oar SI, IffC..y) is eontinuogs for all x # O and y #0. and $(0,0) = O.then tim, Fy) = 0. 52, If F(x) and g() are continuous functions of x, andy, respectively, then Ne) =F) +80) is a continuous funetion of x and y eGo Use the €-8 definition of limit to verify the Ht stazemens given in Problems 53-56, 53. lim, x? +3y7) = 0 to 100) 584. tim te eco oan + 55. tim ~—* <0 toto FS 56, tim | 0, then there exists a 5-neighborhood about (a, such that F{e,y) > O for every point (x,y) in the neighborhood, rove the scalar multiple rule a tim lim LaFles=ar fim, eco ann) Foy) Prove the sum rule: lim Uf + aleey =LiM where L = | im fcs,y) and Ms lo Bet yy) ‘A function of two variables f(x.) may be continuous in each separate variable at.x = x9 and y = yo without being itself continuous at (40, 0). Let/ x43) be defined by ay Soe re #0 0 for (0,0) fouy) Let g(x) = F(+,0) and f(y) = (0,3). Show that both .¢(5) and f(y) are continuous at 0 bor that f(x,y) fs not continuous at (0,0), IN THIS SECTION: Partial differentiation, partial derivative as a slope, partial derivative as arate, higher-order partial derivatives In many problems involving functions of several variables, the goal is to find the derivative of the funetion with respect to one of its variables when all the others are held constant. In this section, we consider this coneept and we will see hhow ican be used to find slopes and rates of change, Partial Differentiation Its often important 10 know how a function of two variables changes with respe B85 Jaw, the pressure ol ‘one of the variables. ‘or example, according to the is related to ils temperature and volume by the formula P constant while the volume is allowed to vary, we might want lo know the effect on the rate of change of pressure. Similarly lowed to vary, We might w Ifthe temperature is kept constant while the temperature i the rate of change of pressure ‘Tho process of differentiating a Funct its variables while keeping the other v and the resulting derivative Recall that the derive limit of « dierence quotient, namely, " Se + 4x) -f) F0)= Sin Partial deri tives with respect to x or y are defined similar. SE where & is constant, if the volume is kept to know the effect on n of several variables with respect to one of iable(s) fixed is called partial differentiation, ‘partial derivative of the function, ‘of a function of a single variable f is defined 10 be the11.3 Partial Derivatives : For the partial differentiation of a function of two variables, 1 (4.9). We find the partial derivative with respect 10.x by regarding y as constant ‘while differentiating the function with respect to x. Similarly, the partial derivative with respect to’ is found by regarding x as constant while differentiating with respect to y. If G,y) = 8ybaty? finds a fe bef Sotution a. For f,, hold y constant and find the derivative with respect to x: f(t) = xy +29? b, For f,, hold x constant and find the derivative with respect :0 y: e009) Several different symbols are used to denote partial derivatives, as indicated in the following box. Ba Duty