Lecture 10

Download as docx, pdf, or txt
Download as docx, pdf, or txt
You are on page 1of 11

Lecture 10

Application of differential calculus to investigation of behaviour of functions.

Monotonicity
Conditions that a function may be constant.
.
THEOREM. Suppose that the function f (x) is defined in the interval I and has
inside it a finite derivative f ( x ) and is continuous at the end-points (if they belong to I ). In
order that f (x) be a constant, it is sufficient that f ´ ( x )=0 inside I .
Proof. Fix a point x 0 ∈ I and take an arbitrary point I ∋ x ≠ x 0.In the interval [ x , x 0 ] or
[ x 0 , x ] all conditions of Lagrange's theorem are satisfied.
Consequently we may write
f ( x )−f ( x 0 )=f ´ (c )(x−x 0) ,
where c being between x and x 0: x 0 <c < x∨x <c < x 0.
But, according to the assumption f ´ ( c ) =0 and therefore for all x from I
we have f ( x )=f ( x 0 )=constant , which proves our proposition.
Observe that it is evident that the above condition is also necessary
for a function to be constant, since for any constant function f ( x ) ≡C , C ' =0.
The following simple corollary has an important application in the integral calculus.
COROLLARY. Suppose that two functions f ( x ) and g( x ) are defined in the same
interval I , inside it have finite derivatives f ´ ( x ) and g ´ ( x ), and the functions are
continuous at the end-points (if they belong to I ).
If, moreover,
f ´ ( x )=g ´ ( x ) inside I ,
then the functions f ( x )∧g ( x )differ by only a constant C over the whole interval I
f ( x )=g ( x ) +C
Definitaion. A function f (x ) is called increasing in an interval (a, b) if for any two

points x 1 , x 2 ∈(a , b )
x 1 < x 2 ⇒ f ( x1 )< f ( x2 )
In other words, a function is increasing if the values of the functions increase as the
argument increases.

y y

f(x2) f(x2)

f(x1) f(x1)

o x2 x o x1 x2 x
x1
A function f (x ) is called decreasing in an interval ( a , b ) if for any two points
x 1 , x 2 ∈(a ,b )
x 1 < x 2 ⇒ f ( x1 )> f ( x2 )
In other words, a function is decreasing if its values decrease as the argument increases.

y y

f(x2)
f(x2)
f(x1) f(x1)

o x2 o x1 x2 x
x1 x

Both increasing and decreasing functions are called monotonic.

Conditions for Monotonicity


Between the character (nature) of the monotonicity of differentiable function and the
sigh of its derivative there is a connection stated in the following theorem.
Theorem. Let f (x ) be differentiable function in an interval (a, b). Then
¿ ¿
i) if f ( x )>0 ⇒ f ↑ increases ⇒ f ( x )≥0 ;
¿
ii) If f ( x )<0 ⇒ f ↓ decreases ⇒ f ¿ ( x )≤0 .
Proof. Let us prove the first statement, the second is proved similarly.
¿ ¿
Suppose f ( x) is positive in an interval (a, b), i.e. f ( x )>0 for all x ∈(a, b ) . Now let us

take arbitrary two points x 1 , x 2 ∈(a ,b ) and assume that x 1 <x 2 . According to the
Lagrange’s theorem we can write
¿
f (x 2 )−f ( x 1 )=f (ξ )( x 2−x 1 ) , x 1 <ξ < x 2
It is clear that
¿ ¿ ¿
if f ( x )>0 for x ∈(a , b )⇒ f ( ξ )>0 ⇒ f (ξ )⋅(x 2 −x 1 ) > 0 ⇒ f ( x 2 )−f ( x 1 )>0 ,

Thus, we have for any two points x 1 , x 2 ∈(a , b ) : x 1 < x 2 ⇒ f ( x1 )< f ( x2 ) .


That is the function f (x ) is increasing in (a, b).
The left part of the first statement has proved. Now let us prove the right part of the
first statement. In other words we have to prove that
f ↑ increases ⇒ f ¿ ( x )≥0
¿
Since f (x ) is differentiable function in (a, b), that is f (x ) possesses the derivative f ( x)
at any point x ∈(a ,b ) , we have
f ( x+ Δx)−f (x )
f ¿ ( x )= lim
Δx → 0 Δx
It is easy to see that for Δx>0 we have
x+ Δx > x ⇒ f ( x + Δx )> f ( x )⇒ f ( x + Δx )−f ( x )>0 ⇒
f ( x+ Δx )−f ( x ) f ( x + Δx )−f ( x ) ¿
⇒ >0 ⇒ lim ≥0⇒ f ( x )≥0
Δx Δx→+0 Δx

and for Δx<0 we have


x+ Δx< x ⇒ f (x + Δx )< f ( x )⇒ f ( x + Δx )−f ( x )<0 ⇒
f ( x+ Δx)−f ( x ) f ( x+ Δx)−f ( x ) ¿
⇒ >0 ⇒ lim ≥0 ⇒ f ( x )≥0
Δx Δx →−0 Δx
¿
Thus, for any case f ( x )≥0 which is what we had to prove.
Example. Find the intervals of increasing and decreasing of the function
4 2
f (x )=x −8 x −3 .
¿ 3
Solution. We find f ( x)=4 x −16 x ¿ 4 x ( x+ 2 )( x−2 ).
It is easy to verify that
¿
f ( x )>0 for x ∈(−2,0)∪(2 ,+∞) and
¿
f ( x )<0 for x ∈(−∞ ,−2 )∪(0,2 )

Thus, f (x ) increases in [ −2,0 ] ∪[2 ,+∞ ) and f (x ) decreases in (−∞ ,−2]∪[ 0,2 ] .

Points of Extremum. Extrema of Functions

Definition. A point x 0 is called a point of local maximum of the function f (x ) if the

value f (x 0 ) is the greatest value of the function f (x ) in a neighbourhood of the point x 0 .


In other words , it means that there exist a δ - neighbourhood, δ > 0 , of the point
x 0 , U δ ( x 0 )=( x 0 −δ , x 0 +δ ) such that for all x ∈ U δ (x 0 ) f ( x )≤f ( x 0 ) .
y y y
max
f(x0) max
f(x0) max f(x0)

We alsoo say that xf (x 0 ) is a local maximum


o of
x0 the function
x
f o(x ) . x0 x
0 x

local
y max
max
max

o x

Here α , β ,γ are points of local maximum of the function f (x ) , and f (α ), f ( β ), f (γ ) are


local maximum of the function f (x ) . The point of local minimum and the local minimum
of a function are defined in a similar way.

Definition. A point x 0 is called a point of local minimum of the function f (x ) , if the

value f (x 0 ) is the least value of the function f (x ) in a neighbourhood of the point x 0 .

We also say that f (x 0 ) is a local minimum of the function f (x ) .


The points of local minimum and maximum denoted as x min and x max and called in
common the extremum points .The value of function in extremum points called
extremum of function.
Assume first that the function f(x) has in the interval ( a , b ) a finite derivative. If at a
point x 0 the function has an extremum, then applying Fermat's theorem to the interval
( x 0−δ , x0 + δ ) ,which we introduced above, we find that f ´ ( x 0 ¿=0 , which is the necessary
condition for an extremum. The extremum should be sought only at the points at which
the derivative vanishes; such points will be called stationary. However, one should not
think that every stationary point provides the function with an extremum; the above
necessary condition is not sufficient. For instance, the derivative 3 x 2 of the function x 3
vanishes for x = 0 but at this point the function has no extremum; it increases all the time.
Thus we can say only that the stationary point of a function f (x) is "suspect" and
should undergo a further investigation. If we extend the class of considered functions to
admit those which have no finite derivative at isolated points, it is possible that the
extremum may occur at one of these points. They should also therefore be classified
among "the suspicious points" and should be investigated.
Definition. A point x 0 is called critical of the function f ( x ) if in this point there isn't
derivative or f ' ( x 0 )=0.
Theorem1.Let f ( x)∈C ( x 0−δ , x 0 +δ ) . If f ´ ( x ) >0(¿ 0) as x ∈(x 0−δ , x 0) and f ´ ( x ) <0(¿ 0) as
x ∈ ( x 0 , x 0 +δ ) , then x 0 is the maximum (minimum) point.

If f ´ ( x ) >0( ¿ 0) as x ∈(x 0−δ , x 0)∪ ( x 0 , x 0 +δ ) then x 0 is not extremum point.


Proof. If f ´ ( x ) >0(¿ 0) as x ∈ ( x 0−δ , x 0 ) , then the function f strictly increasing
(decreasing) in ( x 0−δ , x0 ) , i.e. f (x)< f ( x 0) as x ∈ ( x 0−δ , x 0 )and f ´ ( x ) <0(¿ 0) as x ∈ ( x 0 , x 0 +δ ) ,
then the function f strictly decreasing (increasing) in ( x 0 , x 0 +δ ), i.e. f (x)< f ( x 0)as x ∈
( x 0 , x 0 +δ ). By virtue of f (x)∈C (x 0−δ , x 0 +δ) we get that x 0 is the maximum (minimum)
point.
If f ´ ( x ) >0( ¿ 0) as x ∈ ( x 0−δ , x 0 ) ∪ ( x 0 , x 0 +δ ),then the function f ( x) increasing (decreasing) in
interval ( x 0−δ , x0 + δ ),i.e. is not extremum point.
Theorem2.Let the function f (x) possesses the first derivative at some neighborhood
N δ (x 0 ) of the stationary point x 0 , i.e. f ´ ( x 0 ¿=0 and besides, there exist non zero second

derivative at this point f ´ ´ ( x 0 ¿ ≠ 0 . Then x 0is the extremum point, More exactly, if f ´ ´ (
x 0 ¿ <0then x 0is the maximum point, and if f ´ ´ ( x 0 ¿ >0then x 0is the minimum point.

Proof. In fact, according to the definition of the derivative, taking into


account f ´ ( x 0 ¿=0 ,we have
f ´ ( x )−f ´ ( x 0) f ´ (x) f ´ (x)
f ´ ´ ( x 0 ¿= lim = lim .Let f ´ ´ ( x 0 ¿ <0 .Then x−x <0 for sufficiently close
x→ x 0 x−x 0 0 x→ x x−x 0 0

to x 0 points x .Therefore for all considered values of x , if x < x 0 , then necessarily f ´ ( x ) >0,
and if x > x 0 , then necessarily f ´ ( x ) <0, By the Theorem1, x 0 is the maximum point.
Similarly we establish that if f ´ ´ ( x 0 ¿ >0then x 0is the minimum point.
Theorem3. Let I be an interval and x 0 be an interior point of I and n ≥ 2. Suppose
that the derivatives f ´ , ´ ´ , … , f (n) exist and are continuous in a some neighborhood N δ ( x 0 )
( n−1 ) (n )
of x 0 that f ´ ( x 0 ¿=f ´ ´ ( x 0 ¿=f ( x 0 )=0 , but f ( x 0 ) ≠ 0.
(n )
A) If n is even and f ( x 0 ) <0, then f has a local maximum at x 0;
(n )
B) If n is even and f ( x 0 ) >0, then f has a local minimum at x 0;
C) If n is odd , then f neither a maximum nor a minimum at x 0.
Proof. Applying Taylor’s theorem at x 0 , for x ∈ I we have
f ( n) (c)
( x−x 0 ) ,
n
f ( x )=f ( x 0 ) +
n!
(n )
where c is some point between x 0and x .Since f (n)is continuous ,if f ( x 0 ) ≠ 0, then
there exists an interval U ( x 0 ) containing x 0 such that f (n ) ( x )will have the same sign as
f (n ) ( x 0 ) for x ∈ U ( x 0) and c ∈U ( x 0 ) and consequently f (n ) ( c ) and f (n ) ( x 0 )will have the same

sign.
(n ) n
If n is even and f ( x 0 ) <0,then f (n ) ( c ) <0 and ( x−x 0 ) ≥ 0 and so that f ( x )−f ( x 0 )≤ 0, or
f ( x ) ≤ f ( x 0 ) .Therefore f has a local maximum at x 0.If f (n ) ( x 0 ) >0 , then f (n ) ( c ) >0and

f ( x ) ≥ f ( x 0 ). Therefore f has a local maximum at x 0.


n
If n is odd, then ( x−x 0 ) positive if x > x 0 and negative if x < x 0 . Consequently , if
x ∈ U ( x 0) , then f ( x )−f ( x 0 ) will have opposite signs to the left and to the right of x 0.

Therefore, f has neither a maximum nor a minimum at x 0.


Convex Functions
Definition. (Convex Function, Concave Function). A function f : I → R is said to be
convex on I if for every 0 ≤ t ≤ 1 and any points x 1 , x 2 ∈ I we have
f ( (1−t ) x 1+ t x 2) ≤ ( 1−t ) f ( x 1 ) +tf (x 2 ) .

A function f : I → R is said to be concave on I if the function −f (x) is convex on I .


Note that if x 1< x2, then as t ranges from 0 to 1,the point ( 1−t ) x1 +t x 2the traverses interval
from x 1 to x 2 .Thus f is convex on I and if x 1 , x 2 ∈ I then the chord joining any two points
A ( x 1 , f (x 1) ) and B ( x2 , f ( x2 ))on the graph of f lies above the graph of f .

A convex function need not be differentiable at every point, as the example


f ( x )=| x|, x ∈ R.reveals.However, it can be shown that if I is an open interval and if f : I → R
convex on I , then the left and right derivatives of f exist at every point of I . As a
consequence, it follows that a convex function on an open interval is necessarily
continuous.
Proposition (Jensen’s Inequality). If f : I → R is convex on I then it satisfies Jensen’s
inequality:

( )
n n
f ∑ λk xk ≤ ∑ λk f ( xk )
k=1 k=1
n

for any x 1 , … , x n ∈ I and any λ 1 , … , λn ∈ [ 0,1 ] , with ∑ λk =1.


k =1

Proof. We use definition and induction, assuming inequality for any x i ∈ I and λ i ∈ [ 0,1 ]
m
with ∑ λi=1and m<n .We then have , for λ i<1 ,
i=1

( ) ( ) ( )
n n−1 n−1 n
λ λ
f ∑ λ k x k =f ( 1− λn ) ∑ 1−λn x j + λ n x n ≤ ( 1− λn ) f ∑ 1−λn x j + λn f ( x n ) ≤ ∑ λ k f ( x k )
k=1 j=1 j j=1 j k=1

Theorem 4 . Let I be on open interval and suppose that f : I → R has a second

derivative on I .Then f is a convex function on I if and only if f ´ ´ ≥ 0 for all x ∈ I .


Infilection point
Consider a function   y=f ( x ), wich is continuous at a point  x 0.The
function  f (x) can have a finite or infinite derivative  f ′( x 0) at this point. If, when
passing through   x 0, the function changes the direction of convexity, i.e. there
exists a number  δ>0 such that the function is convex upward on one of the
intervals ( x 0−δ , x0 )  or ( x 0 , x 0 +δ ), and is convex downward on the other, then  x 0 is
called a point of inflection of the function  y=f ( x ).
The geometric meaning of an inflection point is that the graph of the function  f ( x) 
passes from one side of the tangent line to the other at this point, i.e. the curve
and the tangent line intersect.
Definition. The point M ( x 0 , f ( x 0 ) ) of the curve is called its inflection point if it
separates the section of the curve where the function f (x) is convex from the
section where this function is concave.

Another interesting feature of an inflection point is that the graph of the


function   f ( x ) in the vicinity of the inflection point  x 0 is located within a pair of the
vertical angles formed by the tangent and normal .

Necessary Condition for an Inflection Point (Second Derivative Test)


If  x 0 is a point of inflection of the function  f ( x ), and this function has a second
derivative in some neighborhood of  x 0, which is continuous at the point  x 0 itself,
then
f ´ ´ ( x 0 ¿=0
Proof. Suppose that the second derivative at the inflection point  x 0 is not zero
  f ´ ´ ( x 0 ¿ ≠ 0.
 Since it is continuous at   x 0, then there exists a δ-neighborhood of the point  x 0 
where the second derivative preserves its sign, that is

. f ´ ´ ( x ¿ <0 or f ´ ´ ( x ¿ >0 ∀ x ∈ ( x x −δ , x 0 +δ )

In this case, the function is either strictly convex upward (when   f ´ ´ ¿)<0) or
strictly convex downward (when f ´ ´ ( x )¿ 0). But then the point   x 0 is not an
inflection point. Hence, the assumption is wrong and the second derivative of the
inflection point must be equal to zero:
f ´ ´ ( x 0 ¿=0 .

First Sufficient Condition for an Inflection Point (Second Derivative Test)


If the function  f ( x) is continuous and differentiable at a point  x 0, has a second
derivative f ´ ´ ( x 0 ¿  in some deleted δ-neighborhood of the point  x 0 and if the
second derivative changes sign when passing through the point  x 0, then  x 0 is a
point of inflection of the function  f (x) .
Proof. Suppose, for example, that the second derivative f ´ ´ (x)  changes sign from
plus to minus when passing through the point  x 0 . Hence, in the left δ-
neighborhood ( x 0−δ , x0 ) the inequality  f ´ ´ ( x ¿ >0  holds, and in the right δ-
neighborhood ( x 0 , x 0 +δ ) , the inequality  f ´ ´ ( x ¿ <0 is valid.
In this case, according to the sufficient conditions for convexity, the function  f ( x
) is convex downward in the left δ-neighborhood of the point  x 0 and is convex
upward in the right δ-neighborhood.
Consequently, the function changes the direction of convexity at the point  x 0, that
is by definition,  x 0 is a point of inflection.

Second Sufficient Condition for an Inflection Point (Third


Derivative Test)
Let  f ´ ´ ( x 0 ¿=0 ,  f ´ ´ ´ ( x 0 ¿ ≠ 0. Then  x 0  is a point of inflection of the function .
Proof.
As   f ´ ´ ´ ( x 0 ¿ ≠ 0, the second derivative is either strictly increasing at  x 0 (if ,  f ´ ´ ´ (
x 0 ¿ >0) or strictly decreasing at this point (if ,  f ´ ´ ´ ( x 0 ¿ <0). Because  f ´ ´ ( x 0 ¿=0 , then
the second derivative for some δ>0 has different signs in the left and right δ-
neighborhood of  x 0. Hence, on the basis of the previous theorem, it follows that  x 0 
is a point of inflection of the function  f ( x).
Asymptotes
An asymptote of a curve  y=f ( x ) that has an infinite branch is called a line
such that the distance between the point ( x , f ( x )) lying on the curve and the line
approaches zero as the point moves along the branch to infinity.
Asymptotes can be vertical, oblique (slant) and horizontal. A horizontal
asymptote is often considered as a special case of an oblique asymptote.
Vertical Asymptote
The straight line  x=a  is a vertical asymptote of the graph of the function
  y=f ( x ) if at least one of the following conditions is true:
lim f ( x )=± ∞ , lim f ( x )=± ∞.
x→ a−0 x → a+0

In other words, at least one of the one-sided limits at the point  x=a must be equal
to infinity.
A vertical asymptote occurs in rational functions at the points when the
denominator is zero and the numerator is not equal to zero (i.e. at the points of
1
discontinuity of the second kind). For example, the graph of the function   y= x
has the vertical asymptote ¿ 0 . In this case, both one-sided limits (from the left
and from the right) tend to infinity:
1 1
lim =−∞ , lim =+∞
x→ 0−0 x x→ 0+0 x

A function which is continuous on the whole set of real numbers has no vertical
asymptotes.
Oblique Asymptote
The straight line  y=kx+ b is called an oblique (slant) asymptote of the graph of
the function  y=f (x ) as  x →+∞ or as x →-∞  if

lim [ f ( x )−( kx +b ) ]=0 or lim [ f ( x )− ( kx+ b ) ] =0


x→+∞ x→−∞

The oblique asymptotes of the graph of the function  y=f ( x ) may be different as  x
→+∞  and   x →−∞.
Therefore, when finding oblique (or horizontal) asymptotes, it is a good practice
to compute them separately.
The coefficients k  and b of an oblique asymptote  y=kx+ b are defined by the
following Theorem:
A straight line  y=kx+ b   is an asymptote of a function
  y=f ( x )  as   x →+∞ if and only if the following two limits are finite:
f (x) lim [ f ( x )−kx ] =b.
lim =k and
x→+∞ x x→+∞

Proof.Necessity
Let the straight line  y=kx+ b be an asymptote of the graph of
  y=f ( x )  as  x →+∞. Then the following condition is true:
lim [ f ( x )−( kx +b ) ] =0
x→+∞

or, equivalently,
lim α ( x )=0 .
f ( x )=kx+ b+α ( x) as x →+ ∞, where x→+∞

Dividing both sides of the equation by   x , we obtain:


f (x ) kx+ b+α (x) b α (x)
= =k + + .
x x x x
Consequently, in the limit as  x →+∞, we have

lim
x→+∞
f (x)
x [
b α( x )
= lim k + +
x →+∞ x x ] and

lim [ f ( x )−kx ] = lim [ b+ α (x ) ] =¿ b ¿.


x→+∞ x →+∞

Sufficiency
Suppose that there are finite limits
f (x)
lim =k∧ lim [ f ( x )−kx ] =b .
x→+∞ x x→+ ∞

The second limit can be written as


lim [ f ( x )−(kx +b) ]=0 ,
x→+∞

that meets the definition of an oblique asymptote. Thus, the straight line
  y=kx+ b is an asymptote of the function  y=f ( x ).
Note:
Similarly we can prove the theorem for the case of  x →−¿∞.

Horizontal Asymptote
In particular, if  k =0, we obtain a horizontal asymptote, which is described by
the equation  y=b. The theorem on necessary and sufficient conditions for the
existence of a horizontal asymptote is stated as follows:
A straight line  y=b is an asymptote of a function  y=f (x ) as  x →+∞, if and only if
the following limit is finite:
lim f ( x )=b
x→+∞

The case  x →−∞ is considered in the same way.

You might also like