Function of many variables. Partial derivatives.

Partial
differentials of function of many variables. Total differential

2.1 Conception of function of many variables

Thus far, we have considered functions of a single variable, e. g.
y(x), f(t). In practical applications, however, quantities of interest are
usually functions of many variables x1, x2, x3,…xn. Then y is called
function of n variables x1, x2, x3,…xn and they write it as y=f(x1, x2,
x3,…xn).
Most real-world functions depend on more than one variable:
- displacement of a violin string: position x and time t;
- steady state temperature in a metal plate: position x and y;
- behavior of an electron in a transistor: position x, y, and z and time t.
For example, consider the ideal gas law
pV = nRT;
where p is the gas pressure; V is the volume of the gas; n is the number of
moles of the gas; R is the molar gas constant; and T is the temperature of the
gas. It follows that the gas pressure, p, is a function of three variables: n, V and
T (R is a constant so it cannot vary).
Notation is used for functions of many variables:
f(x; y) is a function of two variables, x and y; and g(x; y; t) is a
function of three variables, x, y and t.
Example: z=f(x,y)=4x-3y+3 is a linear function of two variables.

2.2 Visualizing functions of many variables

A function of one variable can be drawn as a curve in two dimensions,
see Fig. 2.1. A function of two variables forms a surface in three-dimensions,
see Fig. 2.2. Unfortunately, functions of more than two variables cannot easily
be drawn in a three-dimensional world! In fact, we shall usually study functions
of two variables only, although the ideas can be extended to functions of more
variables.

Fig. 2.1 Functions of a single variable: f ( x)  x 2 and f ( x)  sin x

 a) b)
Fig. 2.2 Functions of two variables:
a) f ( x, y)  xy and b) f ( x)  cos x sin y

2.3 Rates of change

The ordinary derivative of a function of one variable, dy(x) , represents
dx
the rate of change of the function, y, with respect to the (only) variable, x.
Equivalently, the ordinary derivative is the gradient (slope) of the line y(x).
Equations that involve such rates of change are known as ordinary differential
equations (ODEs). A natural question now arises: for functions of more than
one variable, how do we measure the rate of change? This question leads to the
concepts of partial and total derivatives.
Note that for functions of more than one variable, there is no ordinary
derivative.

2.3.1 Partial derivatives

Definition. The partial derivative of a function of many variables is defined to
be the rate of change of the function with respect to one of the
variables, assuming that all other variables remain fixed.
In practice, this means that partial derivatives can be determined in
exactly the same way as ordinary derivatives by simply treating the
fixed variables as constants.
Notation
If f(x; y) is a function of two variables, then
f or f  is the partial derivative of f with respect to x:
x
x
f or f  is the partial derivative of f with respect to y:
y
y
One can find higher derivatives:

2
 2 f   f  2 f   f 
   , or f   , or f yx
xy x  y  yx y  x 
xy

Note that the symbol  is different to ordinary differentiation d,

and small changes  !
are equivalent: differentiate with respect to y first, then x.

Algebraically definition
The formal definition of the partial derivative is
f ( x, y) f ( x   , y)  f ( x, y )
 lim
x  0 
Note that this is essentially the same definition as the ordinary derivative.

Important Note:
If we have a function of only one variable, f(x), then a common (slight)
f df
abuse of notation is to interpret as the ordinary derivative .
x dx

df ( x) f ( x   )  f ( x)
 lim
dx  0 
2.3.2 Examples
Find the partial derivatives of the function
1. f ( x)  3x; 
 3x   3;
df
dx
2. f ( x, y)  yx; f f
 f ( x, y)x  ( yx )x  y( x)x  y,  x;
x y
3. f ( x, y, z )  sin(xyz ); Let u  xyz , so that sin(xyz )  sin u
sin u   cos u  u

f x  sin( xyz ) x  cos( xyz )  ( xyz )x 
 cos( xyz )  yz ( x)x  yz cos( xyz );
by analogy
f y  xz cos(xyz );
f z  xy cos(xyz );

3
 x
4. z  ln tan
y

x
Let u ( x, y )  , v(u)  tan u , f (v)  ln v ,
y

z  f (v(u( x, y)))


z    f (v(u( x, y)))   f (v)  v(u)  u( x, y)


x  1 1 x  1  1 
u x      ( x)x  , u y     x    y  x    2 .
 y x y y  y x  y  y 

1
v(u )   u ,
cos 2 u

1 1 1
f (v)   v    u .
v tan u cos 2 u

Regarding y as a constant, we get

z x 1 x 1 1 x
 (ln tan )x   (tan )x    ( )x 
x y tan
x y tan
x
cos 2 x y
y y y
1 1 1 2
    .
x x y y sin 2 x
tan cos 2
y y y

Similarly, keeping x constant, we have

z 1 1 x 2x
  ( 2 )   .
y tan x cos 2 x y y 2 sin
2x
y y y

5. u  x3 y 2 z  2 x  3 y  z  5.

4
 u
 ( x 3 y 2 z  2 x  3 y  z  5)x  ( x 3 y 2 z )x  (2 x)x  (3 y )x  ( z )x  (5)x 
x
 y 2 z ( x 3 )x  2  0  0  0  3x 2 y 2 z  2,

Similarly,

u
 2 x3 yz  3,
y
u
 x3 y 2  1.
z

2.4 Total Differential of a Function

1. Total increment of a function:

The total increment of a function z=f(x, y) is the difference

z  f ( x, y)  f ( x  x, y  y)  f ( x, y)

2. The total differential of a function:

The total (or exact) differential of a function z = f(x, y) is the

principal part of the total increment z, which is linear with respect to
the increments in the arguments x and y.
The difference between the total increment and the total
differential of a function is an infinitesimal of higher order compared
with

  x 2  y 2

A function definitely has a total differential if its partial derivatives

are continuous. If a function has a total differential, then it is called
differentiable. The differentials of independent variables coincide with
their increments, that is
5
 dx = x and dy = y.

The total differential of the function z = f(x, y) is computed from the

formula

z z
dz  dx  dy
x y

Similarly, the total differential of a function of three arguments u=f(x,y,z)

is computed from

u u u
du  dx  dy  dz
x y z

2.4.1 Examples
Example 1. Find for the function f ( x, y)  x 2  xy  y 2 the total increment
and total differential.
Solution: Value of the given function in the point ( x  x, y  y) is:

f ( x  x, y  y)  ( x  x)2  ( x  x)( y  y)  ( y  y)2 ;

According to the definition of total increment

f ( x, y )  f ( x  x, y  y )  f ( x, y ) 
 [( x  x) 2  ( x  x)( y  y )  ( y  y ) 2 ]  ( x 2  xy  y 2 ) 
 2 xx  x 2  xy  yx  xy  2 yy  y 2 
 [( 2 x  y )x  ( x  2 y )y ]  (x 2  xy  y 2 ).

Here, the expression (x 2  xy  y 2 ) is an infinitesimal of higher

order compared with x 2  y 2 .

Total differential is

6
 df  (2 x  y)x  ( x  2 y)y.

According to the definition of total differential

f
 ( x 2  xy  y 2 )x  2 x  y;
x
f
 ( x 2  xy  y 2 )y  x  2 y;
y

df  (2x  y)dx  ( x  2 y)dy.

Example 2. Find the total differential of the function

z  x2  y2 .

Solution:

z 1 x z 1 y
  ( x 2 )x  ;   ( y 2 )y  ;
x 2 x 2  y 2 x2  y 2 y 2 x 2  y 2 x2  y 2

x y xdx  ydy
dz  dx  dy  .
x2  y 2 x2  y 2 x2  y2

2.5 Applying the total differential of a function to approximate

calculations
For sufficiently small |x| and |y|, and hence for sufficiently
small

  x 2  y 2

we have for a differentiable function z = f(x, y) the approximate equality

z  dz or

7
 z z
z  x  y.
x y

Example 1.Compute 1.023.01 approximately.

Solution: Consider the function z = xy. The required number may be

considered to be the increased value of this function when x = 1, y = 3, x =
0.02, y = 0.0l. The initial value of the function is z = 1³,

z  dz  yx y 1x  x y ln xy  3 1 0.02  1 ln1 0.01  0.06

Hence,

1.023.01  1  0.06  1.06

Example 2. What is the change in volume of an ideal gas if pressure

increases from 1 to 1.2 atmospheres, and temperature decreases from 300 to 270
K?
Solution:
 The ideal gas law states that
nRT
pV  nRT  V  .
p
 The total change in volume is given by
V V nRT nR nRT dp nRT dT
dV  dp  dT   2 dp  dT    .
p T p p p p p T
Hence the relative change in volume is
dV dp dT 0.2 30
     0.2  0.1  0.3.
V p T 1 300
 The volume decreases by 30 %.