Академический Документы
Профессиональный Документы
Культура Документы
LECTURE 7
Lesson 9
9.1 Partial Derivatives
9.1.1 Calculations
9.1.2 Functions of More Than Two Variables
9.1.3 The Chain Rule
9.1.4 The Chain Rule for Functions of Two Variables
9.1.5 The Chain Rule for Functions of Three Variables
Problems
Lesson 10
10.1 Maxima, Minima, and Saddle Points
10.2 Differential Equations
10.2.1 First Order Differential Equations of First Degree
10.2.1.1 Separable Equations
10.2.1.2 Homogeneous Equations
10.2.1.3 Linear Equations
10.2.1.4 Exact Equations
Problems
------------------------------------------------------------------------------------------
2
Lesson 9
9.1 Partial Derivatives
Partial derivatives are the derivatives we get when all independent variables except one in a
function are kept constant and differentiate with respect to that one.
The Definition
If x0 , y0 is a point in the domain of a function z f x, y , the plane y y0 will cut the
in this plane or the height above (below) the x y -plane is z. The horizontal coordinate is x.
Fig. 9.1(a)
The derivative of z f x, y0 with respect to x at x x0 is defined as the limit (if exists)
d
f f x, y0 f x0 x, y0 f x0 , y0
x0 , y0 d x lim . (9.1)
x x 0 x
at x x0
The limit is called the partial derivative of f with respect to x at the point ( x0 , y0 ) and it is the
to the curve z f x, y0 at x x0 is the line in the plane y y0 that passes through the point
follows.
f Partial derivative of f with respect to x at x0 , y0 .
x0 , y0 or f x x0 , y0
x
9.1.1 Calculations
The definition of the partial derivative f y of z f x, y with respect to y at x0 , y0 is
This limit (if exists) is also taken to be the slope of the curve z f x0 , y in the plane x x0 .
The tangent to the curve is the line in the plane x x0 that passes through x0 , y0 , f x0 , y0
with this slope (see Fig. 9.1(b)).
4
As shown in Fig. 9.1(c), now there are two tangent lines associated with z f x, y at x0 , y0 .
If f x and f y are both continuous at x0 , y0 , the plane determined by these two lines
f x, y x2 3x y y 1 .
f 2
x
x
x 3 x y y 1 2 x 3 y 0 0 2 x 3 y
The value of f x at 4,5 is 2 4 3 5 23 //
2y
EXAMPLE 4 Find f x and f y if f x, y .
y cos x
y cos x 0 2 y sin x
2 y sin x
//
y cos x y cos x
2 2
y cos x 2 2 y 1
2cos x
//
y cos x
2
y cos x
2
4
The definitions of the partial derivatives of functions of more than two independent variables are
similar to the definitions for functions of two variables. They are ordinary derivatives with
respect to one variable, taken while all the other independent variables regarded as constants.
EXAMPLE 5 Find
z
x sin y 3z .
Solution
z
x sin y 3z x sin y 3z x cos y 3z y 3z
z z
3x cos y 3z //
z cos . Find R .
R x y z
Solution i j k sin sin i sin cos j //
1 1 1 1
resistance R given by . Find R R2 .
R R1 R2 R3
Solution Treat R1 and R2 as constants and differentiate both sides of the equation with
2
1 R 1 R R
respect to R2 . Then 2 2 //
R R2 R2 R2 R2
x x(t ), y y (t ), z z (t ) ,
5
we may think of f as a function of the single variable t. At each t, the quantity f at the point
x, y, z on the path is the value of the composite f x(t ), y(t ), z(t ) . If we then wish to know
the rate at which f changes with respect to t along the path, we have only to differentiate this
composite with respect to t and obtain d f (t ) dt . When the formulas are too complicated for
convenient substitution, to find a functions derivatives, we have to use the Chain Rule.
EXAMPLE 8 Use the Chain Rule to find the derivative of f x, y x y with respect to t
along the path x cos t , y sin t . What is the value of the derivative at t 2 ?
Solution We evaluate the right-handed side of Eq. (9.4) with f x, y x y , x cos t , and
y sin t :
f f dx dy
y sin t , x cos t , sin t , cos t
x y dt dt
df f dx f dy
sin t sin t cos t cos t sin 2 t cos2 t cos 2t //
dt x dt y dt
In this case we can check our result by a more direct calculation. As a function of t ,
1 df d 1
f x, y x y cos t sin t sin 2t sin 2t cos 2t //
2 dt dt 2
6
df f dx f dy f dz
. (9.5)
dt x dt y dt z dt
y sin t , z t .
Solution The derivative d f dt shows how the values of f vary along a helix in its
domain. From Eq. (3) we have
df f dx f dy f dz
y sin t x cos t 11 1 cos2t
dt x dt y dt z dt
Definitions
Suppose D is a collection of real numbers x1 , x2 , . . . . . , xn . A real valued function f with
function V r 2 h , the dependent variable is V and the two independent variables are r and h.
Domains
In defining functions of more than one variable, we follow the usual practice of excluding inputs
that lead to complex numbers or division by zero.
7
Solution According to the convention, the domain D is the set of all number pairs ( x, y )
in the plane for which y x 2 is real. This means the points for which
Fig. 9.4:
EXAMPLE 11 Graph the function f ( x, y) 100 x 2 y 2 in space and plot the level curves
f ( x, y ) 0 , f ( x, y ) 51 and f ( x, y ) 75 in the domain of f in the plane.
Solution The domain of f is the entire x y -plane and the range of f is the set of real
numbers less than or equal to 100. The graph is the paraboloid z 100 x2 y 2 ,
a portion of which is shown in Fig. 9.5.
8
Fig. 9.5:
The level curve f ( x, y ) 0 is the set of points in the x y -plane at which
Level Surfaces
In the plane, the points at which a function of two independent variables takes on a constant
value f ( x, y ) c is called a level curve of the function. In space, the points at which a function
of three independent variables takes on a constant value f ( x, y, z ) c make up a surface called
a level surface of the function.
Fig. 9.6:
9
Solution The value of f is the distance from the origin to the point ( x, y, z ) . Each level
(See Fig. 9.6). The level surface x 2 y 2 z 2 0 is not really a surface but
consists of the origin alone.
The tangent plane to the surface z f x, y at the point x0 , y0 , f x0 , y0 lies close to the
surface near the point of tangency. For points x0 h, y0 k near x0 , y0 , therefore, the z -
values along the tangent plane give good approximations to the z -values on the surface. Hence,
to approximate the values of f near x0 , y0 , we can replace the formula for f by the formula
L x, y f x0 , y0 f x x0 , y0 x x0 f y x0 , y0 y y0 . (9.6)
The approximation
L x, y f x, y . . (9.7)
is the standard linear approximation of f near x0 , y0 .
1
EXAMPLE 13 Show that f x, y 1 x y near 0, 0 .
1 x y
1 1
f x x, y , f y x, y .
1 x y 2
1 x y 2
10
L x, y f 0, 0 f x 0, 0 x f y 0, 0 y 1 1 x 1 y 1 x y
The error that may occur in the standard linear approximation of f x, y near x0 , y0 is
defined by E x, y f x, y L x, y and a reasonable estimate of E x, y can be given by
the inequality
M x x0 y y0
1
E x, y
2
, . (9.8)
2
where M is any upper bound for the values of f x x , f y y , and f x y on the rectangular region
1 2
EXAMPLE 15 Find the linearization L x, y of f x, y x 2 x y y 3 near the point
2
Express the upper bound for the error as a percentage of f 3, 2 , the value of f
Solution With f 3, 2 8, f x 3, 2 2 x y 3, 2 4, f y 3, 2 x y 3, 2 1,
M x x0 y y0
1
E x, y
2
Eq. (9.8):
2
To find a suitable value for M , calculate
fxx 2 2 , fy y 1 1, fx y 1 1 .
11
2 x 3 y 2 x3 y2
1
E x, y
2 2
through R.
2
Since x 3 0.1 and y 2 0.1 on R ,
0.04
As a percentage of f 3, 2 8 , the error is no greater than 100 = 0.5% //
8
want to predict how much this value will change if we move nearby to x0 h, y0 k (Fig.
Fig. 9.9:
If h and k are small, f and its linearization at x0 , y0 will change by approximately the same
amount. Since the values of L are always easy to calculate, calculating the change in L gives a
practical way to estimate the corresponding change in f.
12
d f f x x0 , y0 dx f y x0 , y0 d y . (9.9)
in which d f denotes the change in the linearization of f resulting from changes dx and d y in
x and y. We call dx and d y the differentials in x and y and d f the corresponding differential
in f.
True Estimate
Absolute change: f df
f df
Relative change:
f x0 , y0 f x0 , y0
f df
Percentage change: 100 100
f x0 , y0 f x0 , y0
EXAMPLE 17 Suppose that the variables r and h of the function V r 2 h change from the
Estimate the resulting absolute, relative, and percentage changes in the values.
to get
dV 0.2 0.2
0.04 .//
V r0 , h0 r02 h0 1 5
2
dr dh
Solution We are told that 100 2 and 100 0.5 .
r h
dV 2 rh d r r 2 d h 2dr dh
Since
V r2 h r h
dV dr dh dr dh
100 2 100 100 2 100 100
V r h r h
2 2 0.5 4.5
We estimate the error in the volume calculation to be at most 4.5%. //
Lesson 10
10.1 Maxima, Minima, and Saddle Points
Functions of two independent variables can have local maximum and minimum values just the
way functions of single variables can (see Fig. 10.1). We can use partial derivatives to find the
local maximum and the minimum values of a continuous function f x, y on a region R in the
xy -plane by taking the steps below. The routine is much the same as the routine for functions of
a single variable. The main difference is that the first and second derivative tests now involve
more derivatives.
14
x2 y 2
Fig. 10.1: The function z (cos x)(cos y )e has many local maxima and minima.
look for points where the surface z f x, y has a horizontal tangent plane. At such
points we then look for a local maximum, a local minimum, or a saddle point (Fig. 10.3).
examine the values of the function at the boundary points in its domain and at any points
where the first partial derivatives f x or f y do not exist.
Fig. 10.3: Classification of the values of f x, y : Saddle points are the three-dimensional
analogues of points of inflection. Local values are often called relative values.
We organize the search for the extreme values assumed by a continuous function f x, y on a
STEP 1: Make a list that includes the points where f has its local maxima and minima and
evaluate f at all the points on the list.
the local maxima and minima of f can occur only at
(i) boundary points of R ,
(ii) interior points of R where f x f y 0 and points where f x or f y fail to exist.
Therefore, to make the list, we find the maximum and minimum values of f on the
boundary of R and find the values of f at the critical points.
STEP 2: If R is closed and bounded, look through the list for the maximum and minimum values
of f. These will be the absolute maximum and minimum values of f on R .
STEP 3: If R is not closed and bounded, try the following second derivative test.
that f will have an extreme value there. However, if f and its first and second partial
derivatives are continuous on R , there is a second derivative test that may identify the
behavior of f at a, b as shown below.
If f x a, b f y a, b 0 , then
Solution The domain of f has no boundary points, and it is the entire plane (Fig. 10.4). The
The only possibility is the origin, where the value of f is zero. Since f is never
negative, we see that zero is an absolute minimum.
Although the second derivative test is not needed, it would have found
f x x f y y f x2y 2 2 0 4 0
2
fxx 2 , fyy 2, fx y 0 and
Solution The function is defined and differentiable for all x and y and its domain has no
boundary points. The function therefore has extreme values only at the points
where f x and f y are simultaneously zero. This leads to
f x y 2 x 2 0, f y x 2 y 2 0, x y 2
Therefore, the point 2, 2 is the only point where f may take on an extreme
The discriminant of f at a, b 2, 2 is
f x x f y y f x2y 2 2 1 4 1 3
2
f 2, 2 8.
Solution Since the function is differentiable everywhere and its domain has no boundary
points (Fig. 10.5), the function can assume extreme values only where
fx y 0 and fy x 0 .
Thus, the origin is the only point where f might have an extreme value. To see
In the falling body example, we solved the equation dv dt 9.8 on the interval t 0 by finding
the general solution v(t ) 9.8 t C . We then solved the falling body problem by using the
initial condition v (0) 0 to select the particular solution v(t ) 9.8 t from the general
solution.
For example,
2 5
d3y d2y y
3 2 2 ex
dx dx x 1
when y and its derivatives are replaced throughout by f x and its corresponding derivatives.
In this section, only the first one of the following topics, First order equations of first degree,
will be treated. Throughout, only ordinary differential equations will be considered.
a. Variable separable
b. Homogeneous
c. Linear
d. Exact differentials
19
A differential equation of first order and first degree can always be put into the form
M d x N d y 0 .. (10.3)
where M and N represent functions of x and y . When these functions satisfy certain
conditions, the solution is relatively easy to find and can be written as
f x, y C .. (10.4)
The special types of first order and first degree equations we consider are designated:
(a) separable, (b) homogeneous, (c) linear, and (d) exact.
M x d x N y d y 0
This is the standard form of a first order separable equation. Once the equation has been put in
standard form, we can find its solution by direct integration of each term, giving us
M x d x N y d y C
EXAMPLE 5 Solve the differential equation
dy
dx
1 y2 ex .
Solution Rewrite the equation, grouping all the y terms with d y and the x terms with d x :
dy dy
exd x or exd x 0
1 y 2
1 y2
Integration of the first part with respect to x and the second part with respect to y
gives the solution
e x tan 1 y C
y tan 1 e x C //
20
After Eq. (10.8) is solved, the solution of the original equation is obtained by replacing v by y x .
EXAMPLE 6
Show that the equation x 2 y 2 d x 2 x yd y 0 is homogeneous and solve it.
1 y / x
2
dy x2 y 2
dx 2x y 2 y / x
1 v2 y
F v , where v
2v x
dx 2v dv
Then Eq. (10.8) becomes 0
x 1 3v 2
The solution of this equation can be written as
ln x
1
3
ln 1 3v 2 C ,
x3 1 3v 2 e3C C
3y2
In terms of y and x, the solution is
x3 1 2 x x 2 3 y 2 C //
x
21
A linear differential equation of first order can always be put into the standard form
dy
P y Q , .. (10.9)
dx
where P and Q are functions of x.
To solve Eq. (10.9), we find a function = () such that if the given equation is multiplied
by , the left side becomes the derivative of the product y .
dy
P y Q . (10.10)
dx
dy d
P y y (10.11)
dx dx
Expanding the right-hand side of Eq. (10.11) and cancelling terms, we obtain
d
P( x) .. (10.12)
dx
d
P ( x ) d x, ln P( x)d x ln C , .. (10.13)
e
P( x )d x
An integrating factor for Eq. (10.9) . (10.14) (Take C 1 )
d 1
y Q( x)
Eq. (10.9) and Eq. (10.10) y Q( x) d x
dx
As P(x) and Q(x) are given functions of x and, is known from Eq. (10.14), y can be found.
dy
EXAMPLE 7 Solve the differential equation y ex .
dx
e e ex
Pd x dx
the integrating factor from the equation
By multiplying both sides of the original equation by this factor, the left side
becomes
ex
dy
dx
ex y
d x
dx
e y
22
1 2x 1 x
By integrating we get ex y e C y e C e x //
2 2
dy
EXAMPLE 8 Solve the equation x 3 y x2 , x 0.
dx
dy 3 3
Solution Equation in standard form, y x, P( x) , Q( x) x.
dx x x
3/ x dx
e
1 1 1 x 1
e3ln x 3ln x
, and y x3 dx x C,
e x3 x3
y x 2 C x3 //
d f M x, y dx N x, y d y. . (10.18)
f f
M x, y , N x, y .
x y
Integrating the first of these with respect to x , holding y constant, we get
23
f x, y x M x, y dx g y , .. (10.20)
where g y represents an unknown function of y that plays the role of an arbitrary constant.
The integral
x M x, y dx
is an ordinary indefinite integral with respect to x with y held constant. The function f
produced by Eq. (10.20) satisfies the condition
f
M x, y for any choice of the function g y ,
x
To see how we can satisfy the second condition, f y N x, y , we differentiate both sides
f g y
M x, y dx N x, y .
y y x y
dg
N x, y M x, y d x. (10.21)
dy y x
The right side of Eq. (10.21) is a function of y alone, and we can integrate it with respect to y
EXAMPLE 9
Show that the equation x 2 y 2 d x 2 xy cos y d y 0 is exact and solve it.
Solution M x, y x2 y 2 , N x, y 2 xy cos y,
M N
2 y, 2 y, M y N x
y x
f f
Next, to find f x, y such that x2 y 2 , 2 xy cos y,
x y
24
we integrate the first equation with respect to x , holding y constant, and get
x3
f x, y x y2 g y .
3
Differentiating both sides of this equation with respect to y and setting the result
dg dg
equal to N x, y , we get 2 x y 2 x y cos y, or cos y.
dy dy
x3
The solution of the differential equation is x y 2 sin y C. //
3
u v du u dv
d
v v2
y dx x d y
x d y y dx x y 2 dx x dx
y2
x x x2
d x dx 0, C. //
y y 2
The same result would be obtained if we wrote the given equation in the form
xy 2
y dx x d y 0
1 x
x dx 2 d y 0,
y y
1 x
which is exact, since x and we can follow the method
y y x y 2
described earlier for the case of an exact differential equation.