MATHS:: Lecture 15 :: Partial Differentiation
                  
				
Chain Rule differentiation 
				  If y is a function of  u ie y = f(u) and u is a function of x ie u = g(x) then y is related to x  through the intermediate function u ie y =f(g(x) )
				  \y is differentiable  with respect to x
				  Furthermore, let y=f(g(x)) and u=g(x), then 
				  
 =
 
                                      
				  There are a number of related results that  also go under the name of "chain rules." For example, if   y=f(u)   u=g(v), and  v=h(x),
				  then                            
 =   
        
  Problem
				  Differentiate the following  with respect to x
- y = (3x2+4)3
 - y = 

 
 Marginal Analysis
				  Let us assume that the total cost C is represented  as a function total output q.                      (i.e) C= f(q).
				  Then marginal cost is denoted by MC=![]()
				  The average cost = ![]()
				  Similarly if U = u(x) is the utility function of the  commodity x then 
				  the marginal  utility  MU = ![]()
				  The total revenue function TR is the product of  quantity demanded Q and the price P per unit of that commodity then TR = Q.P =  f(Q)
				  Then the marginal revenue denoted by MR is given by ![]()
				  The average revenue = ![]()
Problem
                  1. If the total cost function is C = Q3 -  3Q2 + 15Q. Find Marginal cost and average cost.
  Solution:
                  MC = ![]()
                  AC = ![]()
                  2. The demand function for a commodity is P= (a -  bQ). Find marginal revenue. 
                  (the demand function is generally known as Average  revenue function). Total revenue TR = P.Q = Q. (a - bQ) and marginal revenue  MR= ![]()
  Growth rate and relative growth rate
                  The  growth of the plant is usually measured in terms of dry mater production and as  denoted by W. Growth is a function of time t and is denoted by W=g(t) it is  called a growth function. Here t is the independent variable and w is the  dependent variable. 
                  The  derivative 
is the growth rate (or) the absolute growth rate gr=
. GR = ![]()
                  The  relative growth rate i.e defined as the absolute growth rate divided by the  total 
                  dry  matter production and is denoted by RGR.
                  i.e  RGR = 
. 
 = ![]()
                  Problem 
- If G = at2+b sin t +5 is the growth function function the growth rate and relative growth rate.
 
            GR = ![]()
                  RGR = 
. 
 
  Implicit Functions
                  If  the variables x and y are related with each other such that f (x, y) = 0 then  it is called Implicit function. A function is said to be explicit when  one variable can be expressed completely in terms of the other variable.
                  For  example,   y = x3 + 2x2  + 3x + 1 is an Explicit function
                  xy2 + 2y  +x = 0  is an implicit function
  Problem
                  For example, the implicit equation xy=1 can  be solved by differentiating implicitly gives 
  
=![]()
  ![]()
  ![]()
                  Implicit differentiation is especially useful  when y’(x)is needed, but it is difficult or inconvenient to solve for y in  terms of x. 
  Example:  Differentiate the following function with respect to x ![]()
![]()
  Solution 
                  So, just differentiate as  normal and tack on an appropriate derivative at each step.  Note as well  that the first term will be a product rule.
  ![]()
![]()
![]()
![]()
  Example:  Find ![]()
![]()
![]()
![]()
 for the following function.
  ![]()
![]()
![]()
![]()
  Solution 
                  In this example we really  are going to need to do implicit differentiation of x and write y as y(x).
  
![]()
![]()
 
   Notice that when we  differentiated the y term we used the chain  rule.  
  Example:
   Find ![]()
![]()
![]()
![]()
 for the following. ![]()
![]()
![]()
![]()
  Solutio
                  First differentiate both  sides with respect to x and notice that the first time on left side will  be a product rule.
  ![]()
![]()
![]()
![]()
 Remember that very time  we differentiate a y we also multiply that term by ![]()
![]()
![]()
![]()
![]()
 since we are just using the chain  rule.  Now solve for the derivative.
                    
![]()
![]()
![]()
The algebra in these can be  quite messy so be careful with that.
                    Example
                  Find ![]()
![]()
![]()
![]()
 for the following ![]()
![]()
![]()
![]()
                  Here we’ve got two product  rules to deal with this time.
  ![]()
![]()
![]()
 
   Notice the derivative  tacked onto the secant.  We differentiated a y to get to that point  and so we needed to tack a derivative on. 
   Now, solve for the  derivative.
  
![]()
![]()
![]()
Logarithmic Differentiation
For some problems, first by taking logarithms and then differentiating,
               it is easier to find  
. Such process is called Logarithmic differentiation.
- If the function appears as a product of many simple functions then by
 
taking  logarithm so that the product is converted into a sum. It is now 
                  easier  to differentiate them.
- If the variable x occurs in the exponent then by taking logarithm it is
 
reduced  to a familiar form to differentiate.
     ExampleBegin(); Example  Differentiate  the function.
                  MPSetEqnAttrs('eq0001','',3,[[100,34,16,-1,-1],[133,45,21,-1,-1],[166,56,26,-1,-1],[],[],[],[415,139,67,-3,-3]])    	MPEquation() ![]()
![]()
![]()
![]()
 MPSetEqnAttrs('eq0001','',3,[[100,34,16,-1,-1],[133,45,21,-1,-1],[166,56,26,-1,-1],[],[],[],[415,139,67,-3,-3]]); SolutionDifferentiating  this function could be done with a product rule and a quotient rule. We  can simplify things somewhat by taking logarithms of both sides.
                  MPSetEqnAttrs('eq0002','',3,[[134,40,17,-1,-1],[177,53,22,-1,-1],[221,67,29,-1,-1],[],[],[],[554,165,70,-3,-3]])    	MPEquation() 
![]()
![]()
![]()
                  MPSetEqnAttrs('eq0002','',3,[[134,40,17,-1,-1],[177,53,22,-1,-1],[221,67,29,-1,-1],[],[],[],[554,165,70,-3,-3]]);     	MPSetEqnAttrs('eq0003','',3,[[190,49,22,-1,-1],[253,67,30,-1,-1],[316,84,37,-1,-1],[],[],[],[792,215,95,-3,-3]])    	MPEquation() 
![]()
![]()
 
                  MPSetEqnAttrs('eq0003','',3,[[190,49,22,-1,-1],[253,67,30,-1,-1],[316,84,37,-1,-1],[],[],[],[792,215,95,-3,-3]]);     	MPSetEqnAttrs('eq0004','',3,[[175,89,42,-1,-1],[233,118,55,-1,-1],[291,148,69,-1,-1],[],[],[],[729,368,173,-3,-3]])    	MPEquation() 
![]()
![]()
![]()
                  ExampleBegin(); Example  Differentiate    	MPSetEqnAttrs('eq0007','',3,[[30,11,3,-1,-1],[39,15,3,-1,-1],[47,17,3,-1,-1],[],[],[],[119,44,9,-3,-3]])    	MPEquation() ![]()
![]()
![]()
![]()
![]()
                  MPSetEqnAttrs('eq0007','',3,[[30,11,3,-1,-1],[39,15,3,-1,-1],[47,17,3,-1,-1],[],[],[],[119,44,9,-3,-3]]); Solution
                  First take the logarithm of both sides as we did in the  first example and use the logarithm properties to simplify things a little.
                  MPSetEqnAttrs('eq0009','',3,[[54,30,13,-1,-1],[73,39,16,-1,-1],[90,48,20,-1,-1],[],[],[],[227,120,52,-3,-3]])    	MPEquation() ![]()
![]()
![]()
                  MPSetEqnAttrs('eq0009','',3,[[54,30,13,-1,-1],[73,39,16,-1,-1],[90,48,20,-1,-1],[],[],[],[227,120,52,-3,-3]]);   Differentiate both sides using implicit differentiation.
                  MPSetEqnAttrs('eq0010','',3,[[121,30,12,-1,-1],[162,39,16,-1,-1],[203,49,20,-1,-1],[],[],[],[507,124,50,-3,-3]])    	MPEquation() ![]()
![]()
![]()
![]()
                  MPSetEqnAttrs('eq0010','',3,[[121,30,12,-1,-1],[162,39,16,-1,-1],[203,49,20,-1,-1],[],[],[],[507,124,50,-3,-3]]);   As with the first example multiply by y and substitute  back in for y.
                  MPSetEqnAttrs('eq0011','',3,[[75,34,14,-1,-1],[99,45,19,-1,-1],[123,57,23,-1,-1],[],[],[],[310,145,61,-3,-3]])    	MPEquation() ![]()
![]()
![]()
![]()
PARAMETRIC FUNCTIONS
Sometimes  variables x and y are expressed in terms of a third variable called parameter. We find 
 without eliminating  the third variable.
                  Let  x = f(t) and y = g(t) then 
                      
 =  ![]()
                  =      
 =  
  
  Problem
1. Find  for the parametric function x =a cos
 , y = b sin
 
                Solution
                    
          
 
                    
 =  ![]()
                  =![]()
                  =![]()
  Inference  of the differentiation
                  Let  y = f(x) be a given function then the first order derivative is 
.
                  The  geometrical meaning of the first order derivative is that it represents the  slope of the curve y = f(x) at x.
                  The  physical meaning of the first order derivative is that it represents the rate  of change  of y with respect to x.
                  PROBLEMS  ON HIGHER ORDER DIFFERENTIATION
                  The rate of change of y with  respect x is denoted by 
and called as the  first order derivative of function y with respect to x.
                  The first order derivative of y with respect to x is  again a function of x, which again be differentiated with respect to x and it  is called second order derivative of y = f(x) and is  denoted by  
 which is equal to  
 
                  In  the similar way higher order differentiation can be defined. Ie. The nth order  derivative of y=f(x) can be obtained by differentiating n-1th  derivative of y=f(x)
where  n= 2,3,4,5….
                  Problem
Find  the first , second and third derivative of 
- y =

 - y = log(a-bx)
 - y = sin (ax+b)
 
Partial  Differentiation
				  So far we considered the function of a single variable y  = f(x) where x is the only independent variable. When the number of independent  variable exceeds one then we call it as the function of several variables.
  Example
				  z =  f(x,y) is the function of two variables x and y , where x and y are independent  variables.
				  U=f(x,y,z)  is the function of three variables x,y and z , where x, y and z  are independent variables.
				  In  all these functions there will be only one dependent variable.
				  Consider  a function z = f(x,y). The partial derivative of z with respect to x denoted  by   
and is obtained by differentiating z with respect to x  keeping y as a constant. Similarly the  partial derivative of z with respect to y denoted by 
and is obtained by differentiating z with respect to y  keeping x as a constant.
  Problem 
				  1.  Differentiate   U = log (ax+by+cz) partially with respect to x, y & z
				  We can also find higher order partial  derivatives for the function z = f(x,y) as follows
				  (i)  The second order partial derivative of z with respect to x denoted as 
 is obtained by  partially differentiating 
 with respect to x.  this is also known as direct second order partial derivative of z with respect  to x.
				  (ii)The  second order partial derivative of z with respect to y denoted as 
 is obtained by  partially differentiating 
 with respect to y this  is also known as direct second order partial derivative of z with respect to y
				  (iii)  The second order partial derivative of z with respect to x and then y denoted  as 
 is obtained by  partially differentiating 
 with respect to y.  this is also known as mixed second order partial derivative of z with respect  to x and then y
				  iv)  The second order partial derivative of z with respect to y and then x denoted  as 
  
 is obtained by  partially differentiating 
 with respect to x.  this is also known as mixed  second order  partial derivative of z with respect to y and then x.
				  In  similar way higher order partial derivatives can be found.
  Problem
				  Find  all possible first and second order partial derivatives of 
				  1) z  = sin(ax +by)
				  2) u  = xy + yz + zx   
  Homogeneous  Function
				  A function in which each term has  the same degree is called a homogeneous function.
  Example
- x2 - 2xy + y2 = 0 ® homogeneous function of degree 2.
 - 3x +4y = 0 ® homogeneous function of degree 1.
 - x3 +3x2y + xy2 – y3= 0 ® homogeneous function of degree 3.
 
To  find the degree of a homogeneous function we proceed as follows.
				  Consider the function f(x,y) replace  x by tx and y by ty if f (tx, ty) = tn f(x, y) then n gives the  degree of the homogeneous function. This result can be extended to any number  of variables.
  Problem
				  Find  the degree of the homogeneous function 
- f(x, y) = x2 –2xy + y2
 
- f(x,y)  = 

 
Euler’s  theorem on homogeneous function
				  If U= f(x,y,z) is a homogeneous  function of degree n in the variables x, y & z then 
 ![]()
  Problem
				  Verify  Euler’s theorem for the following function 
				  1.  u(x,y) = x2 –2xy + y2
				  2.  u(x,y) = x3 + y3+ z3–3xyz 
				  Increasing and decreasing function 
  Increasing  function 
				  A  function y= f(x) is said to be an increasing function if f(x1) <  f(x2) for all x1 < x2.
The condition for the  function to be increasing is that its first order derivative is always 
				  greater than zero .
				  i.e    
 >0
  Decreasing  function 
				  A  function y= f(x) is said to be a decreasing   function if f(x1) > f(x2) for all x1  < x2.
The condition for the  function to be decreasing is that its first order derivative is always 
				  less than zero .
				  i.e     
 < 0
Problems
1. Show that the function y  = x3 + x is increasing for all x.
				  2. Find for what values of x  is the function y = 8 + 2x – x2 is increasing or decreasing ?
Maxima and Minima Function of a single variable
         A function y = f(x) is said to have  maximum at x = a if f(a) > f(x) in the neighborhood of the point  x = a and f(a) is the maximum value of f(x) .  The point x = a is also known as local maximum point. 
				  A function y = f(x) is said to have  minimum at x = a if f(a) < f(x) in the neighborhood of the point  x = a and f(a) is the minimum value of f(x) .  The point x = a is also known as local minimum point.
				  The  points at which the function attains maximum or minimum are called the turning  points or stationary points
				  A function y=f(x) can have more than  one maximum or minimum point.
				  Maximum  of all the maximum points is called Global maximum and   minimum of all the minimum points is called Global  minimum.
				  A  point at which neither maximum nor minimum is called Saddle point.
				  [Consider  a function y = f(x). If the function increases upto a particular point x = a  and then decreases it is said to have a maximum at x = a. If the function  decreases upto a point x = b and then increases it is said to have a minimum at  a point x=b.]
				  The necessary and the sufficient condition  for the function y=f(x) to have a maximum or minimum can be tabulated as  follows
  | 
                    Maximum  | 
                    Minimum  | 
                  
First order or necessary condition  | 
                    
  | 
                    
  | 
                  
Second order or sufficient condition  | 
                    
  | 
                    
  | 
                  
Working Procedure
1. Find 
 and 
				2. Equate 
=0 and solve for x. this  will give the turning points of the function.
				3. Consider a turning point  x = a then substitute this value of x in 
  and find the
				  nature of the second derivative.  If  
< 0, then the function has a  maximum 
				  value at the point x = a. If  
> 0, then the function has a  minimum value at 
				  the point x = a. 
				4. Then substitute x = a in the function y = f(x) that will  give the maximum or minimum 
				  value of the function at x = a.
				Problem
Find the maximum and minimum values of the following function
y = x3 – 3x +1
| Download this lecture as PDF here |