AASTU APPLIED MATHEMATICS II

CHAPTER FOUR
DIFFERENTIAL CALCULUS OF FUNCTIONS OF SEVERAL VARIABLES

1: DEFINITIONS AND EXAMPLES OF FUNCTION OF SEVERAL VARIABLES

Definition(Function of two variables)

Let D be the non-empty subset of -plane. A function of two variables x and y is a rule
that assigns a unique real number f ( x, y ) to each point ( x, y ) in some subset D of R2.

Definition(Function of three variables)

Let D be the non-empty subset of three-dimensional spaces. A function of three variables

x , y and z is a rule that assigns a unique real number f  x, y , z  to each point x, y, z  in
some subset D of R3.
Note that:
1. The domain of a function of several variables is the set of all possible values of the variables
for which the function is defined.
2. The range of f is set of all values assumed by f
i.e. Rf = {z / z=f(x, y) form some (x, y) in D}.
Example1: Find the domain of the following
1. f ( x, y)  1  x 2  y 2 , domain of f is {( x, y ) : x 2  y 2  1}.

, domain of f is ( x, y) : ( x, y)  (0,0)
xy
2. f ( x, y )  2
x  y2
3.   
f ( x, y)  y  1  ln x 2  y domain of f is x, y   R 2 /  1  y  x 2 . 
, domain of f is ( x, y) : xy  0 .
1
4. f ( x, y ) 
xy
Example 2: Find the domain of f ( x, y, z )  1  x 2  y 2  z 2 .
 
The domain of f is given by ( x, y, z ) : x 2  y 2  z 2  1 Which is a ball (a solid sphere) of radius 1

2. graph of the function of two variables.

The graph of f  x, y  is the set of all points in R3 given


Graph of f = x, y, z  / z  f x, y  , x, y   D f . 
The graph of functions of two variables is a Surface that can be sketched in R3.
Example 3: Sketch the graph of f ( x, y)  1  x 2  y 2 .
Solution: z  1  x 2  y 2  z 2  x 2  y 2  1 , i.e. the graph of is the set of points in R 3
which are located 1 unit from (0,0,0) and above the xy plane. So it is a hemisphere.

Page 1
 AASTU APPLIED MATHEMATICS II

Graph of

Domain of =

Figure 3.1,
hemisphere

y
Example 4: Sketch the graph of f ( x, y )  1  x  .
2
y y
Solution: Let z  f ( x, y) .Then z  1  x  x  z  1 .We can see that the intersection of the
2 2
1
graph of with the xy plane is a line x  y  1 . Moreover, the intersection of the graph of and
2
xz plane is a line x  z  1. Finally, the intersection of the graph of with the plane is
, i.e. we have the following graph.
z

2
1
x
y

Example 5: Sketch the graph of z  x2  y2

Solution: Clearly, z  0 .Then, the graph is expected to be above the xy plane. Now, we find the
intersection of the graph of f ( x, y)  x 2  y 2 with the planes z  0, y  0 , and x  0. Then,
(i) the intersection of the graph with z  0 ( xy plane) is the point (0, 0).
(ii) The intersection of the graph with is two intersecting lines z  y and
z   y . The intersection of the graph with y  0 is z  x and z   x
.Moreover, for any , z  k intersects the graph z  x 2  y 2 ,i.e.
x 2  y 2  k 2 at a circle of radius r  k with center (0, 0). Thus, contour lines are
circles of radius r. On the other hand, the intersection of the graph of
f ( x, y)  x 2  y 2 with the coordinate axes is intersecting lines passing through
the point .

Page 2
 AASTU APPLIED MATHEMATICS II

Graph of
Contour
map

Example 6: Sketch the graph of z  x 2  y 2 .

Solution: The intersection of the graph z  x 2  y 2 and the coordinate planes xz, yz and xy planes
are the parabola z  x 2 , z  y 2 and the point (0,0) respectively. The contour map for any k  0 is
either a point or a circle. Moreover, since the value of z  0, the graph is above the xy plane. Hence,
we obtain the graph as follows.

Graph of
Contour map

Circular paraboloid

3. LEVEL CURVES
One method of visualizing the function of two variables is looking at its graph. Another method,
borrowed from map markers, is a Contour map on which points of a constant elevation are joined to
form a level curve (contour curve).
Suppose z  f ( x, y ) is a function of two variables and let z  c , c  R be a plane. Then, the
intersection of the graph of z  f ( x, y ) and the plane can be seen to be f ( x, y )  c . One can see
that this is obtained by cutting the graph of z  f ( x, y ) by the plane z  c . Then, the resulting map
(intersection) is called the contour map corresponding to z  c . The projection of this contour map
on the xy plane is called a level curve (contour curve) with a constant. k .
It is generally hard to draw a picture of the graph of a function of two variables. Sometimes, But, we
can sketch level curves. Take a function f of two variables x and y, defined on a domain D. For a real
number k, the level curve at k, denoted C (k ) is the set of all points in D such that f(x, y) = k, i.e.
c(k )  {( x, y )  D : f ( x, y )  k} . The following figures illustrate these two terms.

Page 3
 AASTU APPLIED MATHEMATICS II

Contour map

The plane

level curve (contour lines)

Figure 3.3

Example 7 Let f ( x, y )  x 2  y 2 .

 If k  0, then c (k ) is empty set.

 c(0) contains a single point, namely the origin.
 If k  0, then c (k ) is a circle centered at the origin and whose radius is equal to k .

Graph of

Contour map

Level curves
Circular paraboloid

4: QUADRIC SURFACES

Page 4
 AASTU APPLIED MATHEMATICS II

Definition Let w  f ( x, y, z ) be a function of three variables. Then, the set of points ( x, y, z ) in

R3 such that f ( x, y, z )  k for appropriately chosen real number k , is called the level surface with
respect to the constant k .

Example 8: the level surface of

a. f x, y, z   x 2  y 2  z 2  x 2  y 2  z 2  k , k  0 is a sphere.
b. f x, y, z   x 2  y 2  x 2  y 2  k , k  0 is a Cylinder.
c. f x, y, z   ax  by  cz  ax  by  cz  k it is a planes.

Definition: surfaces defined by polynomials of the form are called Quadric surfaces.

A: ELLIPSOID
The surface whose equation is where a, b, and c are given positive is called an
ellipsoid. Its intersection with any plane, parallel to coordinate planes, is an ellipse. If a  b  c,
then we see that the ellipsoid is a sphere ( see the figure below)

Example 9: a) x 2  4 y 2  z 2  4

b) 9 x 2  4 y 2  z 2  1

B: ELLIPTICAL CYLINDER

2 2
The surface with equation x  y  1 is called an elliptic cylinder. If in particular, a  b, we get
2 2
a b

x  y  a which is called a circular cylinder.

2 2 2

Example 10: y 2  9 z 2  9
C: ELLIPTICAL PARABOLOID

Page 5
 AASTU APPLIED MATHEMATICS II

The surface whose equation is z  ax 2  by 2 where a and b are given positive

numbers, is called a paraboloid.
Example 11: x 2  4 y 2  z  0.
D: HYPERBOLIC PARABOLOID
2 2
The surface whose equation is x  y  z where a, b, c are given positive numbers, is
2 2
a b c
called a hyperbolic paraboloid.

Example 12 z  2 x 2  y 2 .
E: ELLIPTICAL DOUBLE CONE

x2 y2 z2
The surface whose equation is   where a, b, c are given positive numbers, is called
a2 b2 c2
double cone.

Example 13 16 x 2  y 2  4 z 2 .

5: LIMIT OF FUNCTIONS OF TWO /THREE/VARIABLES

Definition: Let f be a function defined throughout a set containing a disk centered at

xo , yo  except possibly at xo , yo  itself, and let L be a real number. L is the limit of f at
 xo , y o  if for every  0 their exists a   0 such that
0 x  xo 2   y  yo 2    f x, y   L   lim f  x, y   L.
 x , y  xo , yo 

Example 14: Show that lim ( x  y )  x0  y 0 using    definition.

( x , y )  ( x0 , y 0 )

Proof:
Let   0 be given. For all ( x, y )  R 2 , we can easily see that
x  x0   x  x 0 2   y  y 0 2 and y  y 0  x  x0 2   y  y 0 2 .Moreover,
we can see that
( x  y )  ( x0  y 0 )  ( x  x0 )  ( y  y 0 )
 x  x0  y  y 0

2  x  x 0  2   y  y 0 2

Choose   . Then, it follows that
2

( x  y)  ( x0  y0 )  ( x  x0 )  ( y  y0 )  x  x0  y  y0  2.  
2
whenever 0   x  x 0  2   y  y 0 2   , ( x , y )  R 2 .
Thus, we conclude that lim ( x  y )  x0  y 0 .
( x , y )  ( x0 , y 0 )

Page 6
 AASTU APPLIED MATHEMATICS II

Theorem.1: The limit of a function of two variables (if exists) is unique.

Theorem 2: Let P  ( x, y ), P0  ( x0 , y 0 ) and suppose lim f ( P)  a, and lim g ( P)  b.
P  P0 P  P0

(a) lim( f  g )( P)  a b (c) lim( f . g ) ( P)  a.b

lP  P0 P  P0

f a
(b) lim(cf ) ( P)  ca
P  P0
(d) lim( g ) ( P)  b , g ( P)  0, b  0.
P  P0

x sin y  x 2  2
Example: 15: Consider the function f ( x, y )  . Then,
x2  y2 1
x sin y  x 2  2
lim f ( x, y )  lim
( x , y ) ( 0 , 0 ) ( x , y ) ( 0 , 0 )
x2  y2 1
lim
( x , y ) ( 0 , 0 )
x sin y  x 2
2 
 (by quotient rule)
lim x2  y2 1
( x , y ) ( 0 , 0 )

 2.
xy
Example 16: Show that lim
( x , y )  (0, 0) x  y2
2 doesn't exist.

xy
Solution: To show that
( x , y )  (0, 0) x  y
2 lim
doesn't exist, it suffice to show that the limits
2

doesn't exist at different direction of approaching the point (0,0) .

Suppose we approach (0,0) through the parabola y  x 2 .Then,
xy x3
lim 2 2 = lim
( x , y )  (0, 0) x  y x0 x2  x4
= 0.

On the other hand, approaching (0, 0) through the line y  x , we get

xy x2 1 1
lim = lim x  lim  .
( x , y )  (0, 0) x  y2
2
x 0
2
x 2
x 0 2 2
xy xy
Hence, since lim  lim , we conclude that
( x , x 2 ) (0, 0) x y
2 2
( x , x ) (0, 0) x  y
2 2

xy
lim
( x , y )  (0, 0) x  y2
2 doesn't exist.

Example 17: Compute the limit at (0, 0) of the function f such that
xy  y 2
f ( x, y )  .
x y
Both the numerator and the denominator approach 0, when (x, y) approaches the origin, thus we
cannot use the quotient rule. We multiply the numerator and the denominator by x  y to obtain

f ( x, y ) 
xy  y 2


( xy  y 2 ) x  y   y( x  y)
x y ( x y ) x y

Page 7
 AASTU APPLIED MATHEMATICS II

Hence, lim f ( x, y)  lim y( x  y )  0.

( x , y ) ( 0 , 0 ) ( x , y ) ( 0 , 0 )

Example 18:

(i) Let f(x, y) =x2y + 2xy3 -1. Then

lim
( x , y )( 2,1)
x 2
y  2 xy 3  1   lim ( x 2 y ) 
( x , y )( 2,1)
lim (2 xy 3  1)  4  3  7
( x , y )( 2,1)

(ii) Let f ( x, y)  x 2  y 2 . Then, lim x 2  y 2  13

( x , y )( 12, 5)

x2  y2
Example 19: Let f ( x, y )  .Then, the domain f is all points in the xy plane except the
x2  y2
origin. We will approach the origin on a line, whose equation is y  mx . Then, we see that
x 2  (mx ) 2 x 2 (1  m 2 ) 1  m 2
f ( x, y )  f ( x, mx )  2  
x  (mx ) 2 x 2 (1  m 2 ) 1  m 2 . If the values m1 and m2 of the variable

m satisfy the condition m1  m2 , then the limits of f ( x, mx ) at 0 are different. Finally, we conclude
that f has no limit at the origin.

x3  y3  1  cosxy 
Example 20: Evaluate a) lim b)  lim 2 2 2 3 
 x , y 0, 0  x 2  y 2
  x , y 0,0  x y  x y 

6: CONTINUITY OF FUNCTIONS OF TWO VARIABLES

Definition: Let f : D  R, where D  R2 be a function and P0  ( x0 , y 0 )  D . Then, f is said to

be continuous at P0  ( x0 , y 0 )

1. Is defined at
2. lim f  x, y  exists
 x , y  xo , yo 
3. lim f  x, y   f  xo , y o 
 x , y  xo , yo 
Theorem 3: Let f and g be continuous at P0 . Then
(i) cf is continuous at P0 , c  R
f  g is continuous at P0 . (iii)
f
(ii) fg is continuous at P0 . (iv) is continuous at P0 . Provided that
g
Theorem 4: Let f be a continuous function of two variables at the point P0 .Then,
(i) If g is a continuous real-valued function at f ( P0 ) , then g  f is continuous.

Page 8
 AASTU APPLIED MATHEMATICS II

(ii) If h is a vector-valued function in R2 which is continuous at t0 and h(t 0 )  P0 , then f  h is

continuous at t 0 .
Theorem 5: Let f be a function of two variables such that f is defined in an open ball centered at
P0  ( x0 , y 0 ) .Then, f is continuous at P0 if and only if

lim f ( P)  f ( P )
P  P0
0

x2  y3
Example 21: The function defined by f ( x, y )  is continuous at the point (0,0) because
x3 y  4
lim f ( x, y )  f (0,0).
( x , y ) ( 0 , 0 )

( x 2  y 2 ) ln( x 2  y 2 ) if ( x, y )  (0,0)
Example 22: Show that f ( x, y )   is continuous at
 0 if ( x, y )  (0,0)
(0,0) .
Solution: Observe that f (0,0)  0 and if (r ,  ) is the polar coordinate of ( x, y ) , we have
x  r cos , y  r sin  .
Therefore, by substitution, we obtain that.
lim ( x  y ) ln( x  y ) = lim r ln(r ), r  x  y
2 2 2 2 2 2 2 2 2

( x , y ) (0, 0) r 0

2 ln r
= lim (by applying L’Hopitals rule)
r0 1
r2
2
 lim r  0
r 0  2

r3
Hence, f is continuous at (0, 0)
Example23: Show that f is not continuous at (0,0) where
 sin ( xy )
 2 if ( x, y )  (0, 0)
f ( x, y )   x  y
2
.
0 if ( x, y )  (0,0)

Solution: Approaching (0,0) through y  x, y   x , we get that
sin( xy ) sin( x 2 ) 1
lim  lim 
( x , y )  (0, 0) x2  y2 x0 2 x2 2
and
sin(  x 2 ) 1
lim 2
 .Therefore the limit of f at (0, 0) doesn’t exist.
x 0 2x 2

7: PARTIAL DERIVATIVES

Definition Let f : D  R2  R be a function of x and y and (x0, y0)  D. Then

Page 9
 AASTU APPLIED MATHEMATICS II

i. The partial derivative of f at (x0, y0) with respect to x is defined as

The partial derivative is denoted by
f ( x0  h, y0 )  f ( x0 , y0 )
f x ( x0 , y0 )  lim (if it exists).
h 0 h
ii. The partial derivative of f at (x0, y0) with respect to y is given by

f ( x0 , y )  f ( x0 , y0 ) f ( x0 , y0  h)  f ( x0 , y0 )
f y ( x0 , y0 )  lim  lim
y  y0 y  y0 h0 h

Example 24: Let f ( x, y )  x 2  y 2 . Find the partial derivatives of f at (0,0).

Solution:
(i) By definition
f ( x, 0)  f (0, 0)
f x (0, 0)  lim = lim x  0.
x 0 x0 x 0

f
Hence, f x (0,0)  x 0 0 . Similarly, f y (0,0)  0 .
x y 0

x y
Example 25: Consider .Then, we see that
x y
f ( x, y )   x  y
  
x x  x  y 
1( x  y )  ( x  y )(1)  2y
  .
( x  y) 2
( x  y) 2
f ( x, y )   x  y 
  
x x  x  y 
1( x  y )  ( x  y )(1) 2y
  .
( x  y) 2
( x  y) 2
xy 2
Example 26: f(x, y) = e cos y. Then,
f ( x, y )
 cos y (e xy ) )( y 2 )  y 2 (cos y )(e xy ).
2 2

x
f ( x, y )
 ( sin y )e xy  2 xy cos ye xy .
2 2

y
= e xy  sin y  2 xy cos y  .
2

Example 27: Find the partial derivatives of the following

a) f x, y   e xy 
x
y

d) f x, y   e x ln x 2  y 2  1 
 y
b) f x, y   9  x 2  y 2 e) f  x, y, z   z sin 1  
x

Page 10
 AASTU APPLIED MATHEMATICS II

f x, y, z   xy
sin z
c)

8: HIGHER ORDER DERIVATIVES

Recall that for a function of two variables and , and are functions of and .
By finding the partial derivatives of , and with respect to and , we obtain second order
partial derivatives as defined below.
(i) The second order partial derivative with respect to the variable is defined by
  f   f 2
  
x  x  x  a x 2 x a
f ( a, b)
y b y b
to denote xx .
(ii) The second order partial derivative with respect to the variable y is defined by
  f  2 f
   2
y  y  x  a y xa
f (a, b)
y b y b
to denote yy .
(iii) The second order mixed partial derivatives at a point are defined by
f x ( a, y )  f x ( a, b) f y ( x, b )  f y ( a , b )
f xy (a, b)  lim f yx (a, b)  lim
y b y b and x a xa provided that
these limits exist. Here, and are denoted by and respectively.
Example 28:
f yx , f xy , f xx , f yy
1. Find
a) f ( x, y)  2 x 2 y 3  3 y 2 x

b)
f ( x, y)  x 2  y 2  1

c) f ( x, y)  Cosh2 ( xy)
d) f ( x, y )  sin( x 2  y 3 )
e) f ( x, y )  3 sin( x  y )  2 cos(x 2 )
f ( x, y)  e x y cos(xy)
2

f)
2. Which of the mixed partial derivatives are equal in question number 1 ? Why?

3. Let f ( x, y )  e x2 y f ,f ,f f .
. Then, find xy yx xyx and yxx
In general, the mixed partial derivatives may not be equal.
f ,f ,f ,f
Theorem 5: Let f be a function of two variables x and y . If x y xy yx are continuous in an
f  f yx
open disk D, then xy at each point of the set D.
Example 29: find f xyy for f x, y   cosxy  x 3  y 4

9: CHAIN RULE
Consider a function of single variable i.e x is a function of a variable t
then, the chain rule states that . In this section, we will see the chain rule for function of
several variables.

Page 11
 AASTU APPLIED MATHEMATICS II

Theorem 6 (First version of the chain rule for two variables)

Let z  f x, y  function of two variables differentiable at ‘x’ and ‘y’ Where
x  g t  and y  ht  is differentiable, then
dz z dx z dy
   
dt x dt y dt

f f
Example 30: Let and x  cos(2t ), y  Cos(t ) Then,  2 x,  3 y 2 . Suppose
x y
dx dy df f dx f dy
now that. Then  2 sin(2t ) ,   sin t . It follows that  .  .
dt dt dt x dt y dt
 2 x(2 sin(2t )  3 y 2 ( cos t )  2 cos(2t )(2 sin(2t )  3 cos2 t ( sin t )

Example 31: Consider f ( x, y )  log(x 2  y 2 ) , where x( )  cos , y ( )  sin  .

df f dx f dy
Then  .  . . But, we can see that
d x d y d
f  1  f  1 
 2 log e  ,  2 2 
2 y log e  .
x  x  y 2
 y  x  y 
dx dy
  sin  ,  cos 
d d
Example 32: Consider f ( x, y )  x 2 cos y , where x(t )  t 2  2t , y (t )  sin t
df f dx f dy
Then,  .  . = (2 x cos y )(2t  2)  ( x 2 sin y )(cost ).
dt x dt y dt
Theorem 7: (First version of the chain rule for three variables):
Let f be a function of three variables x, y, z, differentiable on an open disk centered disk D. Suppose
that x, y and z are functions of a single variable t, differentiable on an open interval I, and such that
for every t  I , x(t ), y(t ), z(t )  D . Then w f (x(t), y(t), z(t) )
df f dx f dy f dz
is a function of t, differentiable on I and  .  .  .
dt x dt y dt z dt

𝜕𝑓
𝜕𝑦
y 𝑦

dw
Example 33: Let w  cos(x z  y)  xy, x(t )  2t , y(t )  t sin t , z (t )  te . Find
2
2 3 2 t
By
dt
applying the chain rule, we compute

Page 12
2,1,1,2. AASTU APPLIED MATHEMATICS II

dy dx dz
 2t sin t  t 2 cos t ,  2,  e t  te t
dt dt dt
w w w
  cos(x 2 z 3  y ).(2 xz 3 ),   cos(x 2 z 3  y ),   cos(x 2 z 3  y ).(3x 2 z 2 )
x y z
dw w dx w dy w dz
Thus by substitution, we get  .  .  .
dt x dt y dt z dt

Theorem 8 (The second version of the chain rule): If x  x(u , v) and y  y (u , v) have first
partial derivatives at (u , v) and z  f ( x, y ) is differentiable at ( x(u , v), y (u, v)) , then
z z x z y
(i)  .  .
u x u y u
z z x z y
(ii)  .  .
v x v y v
Example 34: Consider f ( x, y)  x 2  y tan x where x  u 2  v 2 , y  u  v
f f
Find ,
u v
f f
Solution ,  2x - y sec 2 x,   tan x
x y

Hence,

f f x f y
 .  .  (2 x  y sec2 x)(2v)  tan x
v x v y v

Figure 4.3
The second version of the chain rule

10: IMPLICIT FUNCTIONS OF THREE VARIABLES

 Suppose y and x are variables such that y is a function of x such that y can’t be explicitly
expressed as a function of x . Suppose f is differentiable function of x and y such that
dy
f ( x, y )  0. Then, we need to find at all points at which y is differentiable.
dx

Page 13
 AASTU APPLIED MATHEMATICS II

f dx f dy
By applying the chain rule, we get .  .  0.
x dx y dx
dy f
 x
dx fy
Example 35: Find y  if x 3  y 3  6 xy
Solution: f x, y   x 3  y 3  6 xy  0
dy f 3x 2  6 y 2 y  x 2
 x  2 
dx fy 3 y  6x y 2  2x
 Assume that the variables x, y and z are related by the equation F(x, y, z) = 0. Assuming
that Fz  0 , there is a version of the implicit function theorem which means, we can in
principle, write z = z(x, y), so we can solve for z. Thus, consider the equation   F(x, y,
z(x, y)) = 0 and differentiating both sides partially with respect to x. We get
 F x F y F z F
 .  .  .  0. and since  0, we get
x x x y x z x z
F
z  Fx
   x  .
x F Fz
z
 F x F y F z
 .  .  .  0.
y x y y y z y
F
 Fy
 z   y  .
y F Fz
z
z z
Example 36: Find , if ( i ) x 2  y 2  z 2  9 (ii) xy  z 3  cos(yz 2 )
x y
Solution: (ii) Let F ( x, y, z )  xy  z 3  cos(yz 2 ) . Then,
F ( x, y, z )  0  xy  z 3  cos(yz 2 ) .
Differentiating both sides of the equation xy  z 3  cos(yz 2 ) with respect to y,
we get
1 z  z 
( x  3z 2 )   sin( yz 2 ). z 2  2 yz  which is equivalent to
2 xy  z 3 y  y 

 3z 2  z
  2 yz sin( yz 2 )    z 2 sin( yz 2 ) 
x
.
 2 xy  z 3  y  3
  2 xy z

Thus, solving for

z
, we get 

z  x  2 z sin( yz ) xy  z
2 2 3
.

y y  
4 yz sin( yz 2 xy  z 3  3z 2

Page 14
 AASTU APPLIED MATHEMATICS II

F
 Fy
One can check the result using the formula z   y  .
y F Fz
z

11: THE GRADIENT AND DIRECTIONAL DERIVATIVES

Definition 5.1.1: Let f be a function of two variables that has partial derivatives at ( x0 , y 0 ) . Then,
the gradient of f at ( x0 , y 0 ) is defined as f ( x0 , y0 )  f x ( x0 , y0 )i  f y ( x0 , y0 ) j
.

Now, we give the definition of directional derivatives for a function of two variables as follows.
Definition 5.1.2: Let f be a function defined on a set containing a disk D centered at ( x0 , y 0 ) and
u  a1i  a2 j be a unit vector. The directional derivative of f at ( x0 , y 0 ) in the direction of u ,
denoted by Du f ( x0 , y 0 ) is defined as
f ( x0  ha1 , y0  ha2 )  f ( x0 , y0 )
Du f ( x0 , y0 )  lim
h 0 h
provided that this limit exists.
 If u  (1, 0) , then
f ( x0  h, y0 )  f ( x0 , y0 )  f x ( x0 , y 0 )
Du f ( x0 , y0 )  lim
h 0 h

 If u  j  (0,1) , then
f ( x0 , y0  h )  f ( x0 , y0 )
Du f ( x0 , y0 )  lim = f y ( x0 , y0 ) .
h 0 h
Thus, f x ( x0 , y 0 ) and f y ( x0 , y0 ) are special kind of directional derivatives in the direction of
i  (1,0) and j  (0,1) respectively.
Example 37
Find the directional derivative of the following functions at the indicated points.
(a) f ( x, y )  2 x 2  3xy  y 2  15 at(1, 1) in the direction of u  1 1 .
i j
2 2

(b) f ( x, y)  x 2  y 2 at (1, 2) in the direction of u  1

i 
2
j
.
3 3
(c) f ( x, y, z )  3x  2 y  4 z at (1,1,2) in the direction of u  i  j .
(d) f ( x, y )  y 2 ln x at(1, 4) in the direction of u  3i + 3j.
(e) f ( x, y )  e x cos y at (0,  ) in the direction of u  5i  2 j
4

Now, the following theorem will show that the directional derivative of a function exists in any
direction provided that the limit exists.
Theorem 10: Suppose f is differentiable at ( x0 , y 0 ) . Then f has directional derivative at any
direction. Hence, Du f ( x0 , y 0 )  f ( x0 , y 0 )  u

Page 15
 AASTU APPLIED MATHEMATICS II

Remark: Let f be a function of three variables x, y and z. Then,

(a) f ( x0 , y0 , z0 )  f x ( x0 , y0 , z0 )i  f y ( x0 , y0 , z0 ) j  f z ( x0 , y0 , z0 )k
(b) For any vector u  a1 j  a 2 j  a3 k ( u is unit vector),
Du f ( x0 , y 0 , z 0 )  f ( x0 , y 0 .z 0 )  u.
Example 38: let f x, y, z   x sin yz, find the directional derivative of ‘f’ at (1, 3, 0) in the
direction of v=i+2j-k.
Solution : Dv f 1,3,0   f 1,3,0
v
v
Theorem 11.: Suppose f is differentiable at ( x0 , y 0 ). Then
(a) f ( x0 , y 0 )  0 Implies Du f ( x0 , y 0 )  0 for any vector u .(No rate of change)
(b) If f ( x0 , y 0 )  0 , then
The directional derivative of f in the direction of f ( x0 , y 0 ) is the largest and its value
is f ( x0 , y0 ) . The maximum rate of change of ‘f’ at (xo, yo) occurred when and

have the same direction.

 The minimum rate of change is - which occur when and have

the opposite direction.
Example 39: in which direction does the function f x, y   xy 2  x 3 y increase the fastest at the
point (1,2)? In which direction it decreases fastest.

Solution:
(Exercise)

12: Total Differential

For a differentials of two variables , we define the differentials and and to be
independent variables; i.e. they can be given any values. Then the differential, also called the total
differential, is defined by

Let . Find the total differential

Solution:
Since
and
So
Exercise: Find the total differential of
a)
b)

13: TANGENT PLANES

In this subtopic, we will find the equation of a tangent plane to the graph of a function z  f ( x, y )
at a given point ( x0 , y 0 , f ( x0 , y 0 )) and to the surface f ( x, y, z )  c where f is a function of and f
x

y. For a function f of one variable, if f is differentiable at c, we know that

y  f (c)  ( x  c) f (c) is defined to be the equation of the tangent line to the graph of f at the

Page 16
 AASTU APPLIED MATHEMATICS II

point (c, f (c )) . Analogously, we define the equation of the tangent plane to the graph of ‘f’ of two
variables as follows. See the following graphical illustrations.

Tangent line to the

curve Tangent lines

Cureve in space
Graph of f

Tangent
plane

Definition: Suppose f is differentiable at P0  ( x0 , y 0 ) .Then, the equation of the tangent plane to

the graph of f at P0 , f ( P0 )  is given by z  f ( P0 )  ( P  P0 )  f ( P0 )
or f ( x0 , y0 )  f x ( x0 , y0 )(x  x0 )  f y ( x0 , y0 )( y  y0 )  z  0
Example 40: find the equation of the tangent plane to the elliptical paraboloid
z  3x 2  2 y 2 at 1,1,5
Solution: then the equation of the tangent plane is

Theorem 12: The equation of the tangent plane to the surface f ( x, y, z )  c at the point
( x0 , y 0 , z 0 ) is given by
f ( x, y, z ) f ( x, y, z ) f ( x, y, z )
( x  x0 )  ( y  y0 )  ( z  z 0 )  0.
x y z
Example 41: Find the equation of the tangent plane to the surface
f ( x, y, z )  x 2  y 2  z  9  c at the point P  (1,2,4).
Solution: We have f (1,2,4)  (2,4,1) , and the equation of the tangent plane is
2( x  1)  4( y  2)  ( z  4)  0 .

[ ]

15: EXTREMA OF FUNCTIONS OF TWO VARIABLES

Definition: Let f be a function of two variables. Then,
(a) f is said to have relative maximum value at ( x0 , y 0 ) if and only if there exists a disk
centered at ( x0 , y 0 ) such that f ( x, y )  f ( x0 , y 0 ) for all ( x, y ) in the interior of the disk.

Page 17
 AASTU APPLIED MATHEMATICS II

(b) f is said to have relative minimum value at ( x0 , y 0 ) if and only if there exists a disk centered at
( x0 , y 0 ) and f ( x, y )  f ( x0 , y 0 ) for all ( x, y ) in the disk.
(c) f is said to have absolute maximum at ( x0 , y 0 ) if and only if f ( x0 , y 0 )  f ( x, y ) for all ( x, y )
in the domain of f .
(d) f is said to have absolute minimum at ( x0 , y 0 ) if and only if f ( x0 , y 0 )  f ( x, y ) for all
( x, y ) in the domain of f .

Theorem 13: If f ( x, y ) is continuous in a closed and bounded subset S of the plane, then f has
both absolute maximum/minimum/value in S.
Theorem 14: If f has relative extremum at ( x0 , y 0 ) and f x ( x0 , y 0 ) and f y ( x0 , y0 ) exist, then
f x ( x0 , y0 )  f y ( x0 , y0 )  0.
Definition: Suppose f is a function of two variables. Let ( x0 , y 0 ) be in the domain of f . Then
( x0 , y 0 ) is called the critical point of f if and only if f x ( x0 , y0 )  f y ( x0 , y0 )  0 or either
f x ( x0 , y0 ) or f y ( x0 , y0 ) doesn’t exist.
Example 42 Let f ( x, y )  xy. Then f x  y and f y  x . We have

f x  f y  0  x  y  0. The function f has a single critical point, namely the origin.

In any disk centered at the origin, the function attain minimum through the line y  x. Moreover,
the function attains maximum at (0,0) through the line y   x.

Hence, (0,0) is a saddle point of f. (See the figure below).

Figure 5.3
Graph of

Example 43: Let f ( x,. y )  x 2 y  x  y . Then, we get

f x ( x, y)  2 xy  1  0 and f y ( x, y)  x 2  1  0
1
Solving these equations, we obtain x  1, y   .The function f has four critical points, namely
2
1  1  1  1
(1, ),   1, ,   1,  and 1,  .
2  2  2   2
Theorem 15: (Second partial derivative test): Let f be a function of two variables with continuous
second-order partial derivatives in some disk centered at the critical point ( x0 , y 0 ) .
Let D  f xx ( x0 , y 0 ) f xy ( x0 , y 0 )
 f xx ( x0 , y 0 ) f yy ( x0 , y 0 )  f xy ( x0 , y 0 ).
2

f yx ( x0 , y 0 ) f yy ( x0 , y 0 )
(a) If D  0 and f xx ( x0 , y 0 )  0 , then f has relative minimum at ( x0 , y 0 )

Page 18
 AASTU APPLIED MATHEMATICS II

(b) If D  0 and f xx ( x0 , y 0 )  0 , then f has relative maximum at ( x0 , y 0 )

(c) If D  0, then f has saddle point at ( x0 , y 0 )
(d) If D  0, no conclusion can be drawn.

Example 44: Consider f ( x, y )  x 2  xy  2 y  2 x  1 . Find all the critical points and determine
the relative maximum and relative minimum values of .
Solution: Let ( x, y ) be a relative maximum/relative minimum/ point of f . Then, we get
f x ( x, y )  0 and f y ( x, y )  0 , i.e. we get the system of equations 2 x  y  2  0 with solution
 x20
( 2,2) .Thus, ( 2,2) is the only critical point of . Now, it is easy to see that
f xx ( x, y)  2, f yy ( x, y)  0, f xy ( x, y)  f yx ( x, y)  1 which implies that D  2 1  1  0 .
1 0
Hence, by the second partial test, we conclude that ( 2,2) is a saddle point of .
Example 45: Locate all the critical points and find all the points of relative maximum and relative
minimum of f ( x, y )  x 3  xy  y 3 .
Solution: Let ( x, y ) be a relative extremum point of ‘f’. Then, we see that
f x ( x, y)  3x 2  y  0, f y ( x, y )  3 y 2  x  0
Thus, solving for y in terms of x and substitution yields 27 x 4  x  0 .This is true if
1
x  0 or x  , i.e. (0,0) and   1  1  are
 , 
the only critical points of f.
3  3 3 
Now, we obtain
1 1  1 1
f xx (0,0)  0, f xx ( , )  2, f yy (0,0)  0, f yy  ,   2, f xy ( x, y )  1  f yx ( x, y )
3 3  3 3 
and D (0,0)  1  0, D  1 ,  1   3  0. By the second partial test, we conclude that (0,0) is a
 3 3 
saddle point and   1  1  is a point of relative maximum
 , 
 3 3 

Example 46: Let f ( x, y )  2 x 3  6 x 2  3 y 2  6 xy . Find and classify the critical points of f. By

considering f (x, 0), or otherwise, show that f does not achieve a global maximum.
Solution: We have f x ( x, y)  6 x 2  12x  6 y and f y ( x, y)  6 y  6 x . Thus critical points
occur when y   x and x 2  x  0 , and so at (0, 0) and (1, - 1). Differentiating again
, , and . Thus the discriminator is D (0,0)  36  0 and
D(1,1)  36  0. We have a local maximum at (0, 0) and a saddle point at (1,1).

Example 47: Find a point on the plane 2 x  3 y  4 z  25 closest to the point (3,2,1).

Solution: Let D  ( x  3) 2  ( y  2) 2  ( z  1) 2

where ( x, y, z ) is on the plane 2 x  3 y  4 z  25 .

Page 19
 AASTU APPLIED MATHEMATICS II

Consider the function defined by f ( x, y, z )  ( x  3) 2  ( y  2) 2  ( z  1) 2 .One can easily see that

D attains its minimum value if and only if f attains its minimum value, i.e. finding the minimum
value of f is the same as finding the minimum value of D. Thus, we try to find the minimum value of
f on the plane 2 x  3 y  4 z  25 . To complete the solution, we solve the following systems of
equations.

f 1 3 29
 0  2( x  3)  ( x  y  )  0 (5.3.1)
x 2 4 4

f 31 3 29 
 0  2( y  2)   x  y    0 (5.3.2)
y 22 4 4 

From (1) and (2), we get

10x  3 y  53  0 (5.3.3)

 6 x  25 y  55  0 (5.3.4)

Thus, solving (5.3.3) and (5.3.4), we get x  5 and y  1 . We know that f has absolute minimum
value which is its relative minimum. Now, we use to determine whether the point (5,1) is a point of
relative minimum or not using the second partial test. It is clear that

5 25 3
f xx (5,1)  , f yy (5,1)  and f xy (5,1)  f yx (5,1)  .
2 8 4


Thus, we obtain f xx (5,1) f yy (5,1)  f xy (5,1) 
2 29

5
 0 and f xx (5,1)  .Therefore, by the
4 2
second partial test, we conclude that the point (5,1) is a point of relative minimum. Since f has
minimum value and (5,1) is the only candidate, we conclude that f attains its minimum at (5,1)
.Thus the closets point on the plane to the point for (3,2,1) is .

16: EXTREME VALUES IN A BOUNDED REGION

Consider a function f that is continuous on a bounded subset S of the xy plane. Then, by
extremum value theorem, f attain its absolute maximum and absolute minimum in S. Thus,
the absolute maximum/minimum/ is obtained either on the boundary of S or in the interior of
S. Now, we led to the following steps to find absolute extremum values of f in a bounded
region S.

Step 1: Find critical points that lie in the interior of S.

Step 2: Obtain the maximum/minimum/ value of on the boundary of S.
Step 3: Compare the values of f obtained in step 1 and step 2.

Page 20
 AASTU APPLIED MATHEMATICS II

Example 48: Find the extreme values of f ( x, y )  x 2  y 2 on S where

 
S  ( x, y ) : x 2  y 2  1 .
Solution: f x ( x, y )  0  x  0 and f y ( x, y)  0  y  0.
Thus, (0, 0) is the only critical point of f in the circular region S . Now, let (x, y) be on the boundary
of R. Then, x 2  y 2  1 , i. e. y   1  x 2 .
Thus we change it into single variable function, g ( x)  2 x 2  1. By applying the second derivative
test for
g ( x)  2 x 2  1, we obtain that g 0  1 is minimum value of f . On the other hand,
f 0,1  1. Thus, comparing the values f (0,0)  0 and f 0,1  1, we conclude that
f (0,0)  0 is absolute maximum of f in S. Moreover, f 0,1  1is minimum value in S.
Example 49: A rectangular box with volume 16 cm3 is made from two kinds of material. The top
and the bottom are made of a material costing 10 cents/ sq.cm and the sides from a material costing 5
cents/sq. cm. Find the dimensions of a box with minimum cost.

𝑦
Figure 5.9
Let v( x, y, z )  xyz  16 .The bottom and the top of the box costs 10(2 xy)  20( xy) . Moreover,
the sides cost 5(2 xz)  10( xz) and 5(2 yz)  10( yz) .
Thus, the cost of the box is given by
C ( x, y, z )  20xy  10xz  10 yz
16
Since xyz  16 , we see that z = . Thus, we get c(x, y) = 20xy + 160  160 .
xy y x
Now, we find the critical points using the 1nd partial derivative .

i.e. c  20 y  160  0 and c  20 x  160  0 . Solving for the variable x and substituting,
x 2
x y 2
y
we get 8 y  y  0 which gives y  0 or y  2 .Thus, we obtain x  2. By using the second partial
4

test, it is possible to check that ( 2,2) is a point of relative minimum. Hence, the least dimension of
the box with minimum cost is x  2, y  2 and z  4.

Page 21
 AASTU APPLIED MATHEMATICS II

Example 50: An open-topped rectangular tank is to be constructed so that the sum of the height and
the perimeter of the base is 30 meters. Find the dimensions which maximize the surface area of the
tank. What is the maximum value of the surface area?
Solution: Let the dimensions of the box be as shown.

Figure 5.10

Let the area of the surface of the material be S. Then , S = 2xh + 2yh + xy,

and since, from our restriction on the base and height,

30 = 2(x + y) + h, we have h = 30 - 2(x + y).
Substituting, we have
S  2( x  y )(30  2( x  y ))  xy = 60(x + y) - 4(x + y)2 + xy,
and for physical reasons, S is defined for x  0, y  0, x  y  15 . A global maximum (which we
are given exists) can only occur on the boundary of the domain of definition of S, or at a critical
S S
point, when   0. On the boundary of the domain of definition of S, we have x = 0 or y = 0
x y
or x + y = 15, in which case h = 0. We are given that we may ignore these cases. On the other hand,
we see that

S S
 8 x  7 y  60  0 and  8 y  7 x  60  0 .
x x

Subtracting gives x = y and so 15x = 60, or x = y = 4. Thus h = 14 and the surface area is S = 16(- 4 -
4 - 7 + 15 + 15) = 240 square meters. Since the function S is continuous function on a closed and
bounded region, f has absolute maximum. Moreover, we have only one point of relative maximum.
Thus the absolute maximum exists at this critical point. Sometimes a function necessarily has an
absolute maximum and absolute minimum in the following case because we have a continuous
function defined on a closed bounded subset of R2. In this case exactly as in the one variable case,
we need only search the boundary and the critical points in the interior, using our ability to find local
maxima.

Example 51: Consider f ( x, y )  xy  y  x  1 .Find absolute maximum/minimum/ of in the

closed disk D  {( x, y ) : x  y  4}. .
2 2

Page 22
 AASTU APPLIED MATHEMATICS II

Solution:
f x ( x, y )  0  y  1 and f y ( x, y)  0  x  1.
Then, (1, -1) is the only critical point of f in the interior of the disk D. One can show that (1, -1) is
a saddle point. Then, we search the maximum/minimum/ on the boundary of D. i.e. on the circle x2
+ y2 = 4. One of the parametric equation of x 2  y 2  4 is x(t )  2 cos t , y (t )  2 sin t ,0  t  2 . .
Thus, by substituting x and y in , we get

f ( x, y )  f (t )  4 sin t cos t  2 sin t  2 cos t  1 . Thus, by the second derivative test, one can
show that f assumes maximum at
 1 3 3 1   1 3 3 1  .

 , 
 and 
 , 
 2 2   2 2  

Moreover, f assumes minimum at. √ √

17: LAGRANGE’S MULTIPLIER

One of the applications of the implicit function theorem is the method of finding extreme values of a
function f on the contour line of another function. In this topic, we introduce the procedure called
Lagrange Multiplier Method.

Theorem: Suppose f and g have continuous derivatives throughout a disk centered at P0  ( x0 , y 0 )

such that g ( P0 )  c and that f ( P0 )  f ( P) for all points on the contour line
( x, y) : g ( x, y)  c. If g ( P0 )  0, then there exists  such that f ( P0 )  g ( P0 ) .

The number  is called the Lagrange’s Multiplier.

Example 52: Find the maxima and minima of the function f ( x, y )  x 2  y 2 on the ellipse
x2  2 y 2  1.
Solution: Let g ( x, y )  x 2  2 y 2 and suppose (x, y) is a point of extreme value of f on
x 2  2 y 2  1 .Then, there exists a real number  such that
f ( x , y )   g ( x , y ) .
This is equivalent to solving
 2 x   (2 x)
 for x, y and  .
2 y   ( 4 y )
Since ( x, y ) is on the ellipse, both x and y can’t be zero at the same time. Hence, we get either
1
x  0,   or y  0,   1 . Since we seek points on the ellipse, x  0 implies y
1 and
2 2
y  0 implies that x  1 . Moreover, we see that f (0, 1 )  1 , f (1,0)  1 . We conclude that
2 2
if f has maximum, then it occurs at (1,0) and if it has minimum, then it occurs at (0, 1
)
.
2

Page 23