!"#$%&'(')'*+%,"--.

/$+%0$1%2'3-$%4&"05'
Subject: Real Analysis (Mathematics) Level: M.Sc. Source: Syyed Gul Shah (Chairman, Department of Mathematics, US Sargodha) Collected & Composed by: Atiq ur Rehman (atiq@mathcity.org), http://www.mathcity.org Introduction In elementary treatment of Integral Calculus the subject of integration is treated as inverse of differentiation. The subject arose in connection with the determination of areas of plane regions and was based on the notion of the limit of a type of sum when the number of terms in the sum tends to infinity and each term tends to zero. In fact the name Integral Calculus has its origin in this process of summation. It was only afterwards that it was seen that the subject of integration can also be viewed from the point of the inverse of differentiation. Partition Let [a, b] be a given interval. A finite set P = {a = x0 , x1 , x2 ,...., xk ,...., xn = b} is said to be a partition of [a, b] which divides it into n such intervals [ x0 , x 1 ] , [ x1, x 2 ] , [ x2 , x 3 ] ,......, [ xn-1, x n ] Each sub-interval is called a component of the partition. Obviously, corresponding to different choices of the points xi we shall have different partition. The maximum of the length of the components is defined as the norm of the partition. Riemann Integral Let f be a real-valued function defined and bounded on [a, b] . Corresponding to each partition P of [a, b] , we put M i = sup f ( x) ( xi-1 x xi ) mi = inf f ( x) ( xi-1 x xi ) We define upper and lower sums as Mi U ( P, f ) = M i Dxi and L ( P, f ) = mi Dxi (i = 1,2,...., n)
i =1 b i =1 n n

mi xi 1 xi

where Dxi = xi - xi-1 and finally

f dx = inf U ( P, f ) . (i) f dx = sup L( P, f ) ..(ii)

a a b

Where the infimum and the supremum are taken over all partitions P of [a, b] . Then

f dx
a

and

f dx
a

are called the upper and lower Riemann Integrals of f

over [a, b] respectively. In case the upper and lower integrals are equal, we say that f is RiemannIntegrable on [a, b] and we write f R , where R denotes the set of Riemann integrable functions.

Riemann-Stieltjes Integral

The common value of (i ) and (ii ) is denoted by

f dx
a

or by

f ( x) dx .
a

Which is known as the Riemann integral of f over [a, b] . Theorem The upper and lower integrals are defined for every bounded function f . Proof Take M and m to be the upper and lower bounds of f ( x) in [a, b] . m f ( x) M (a x b ) Then M i M and mi m (i = 1,2,....., n) Where M i and mi denote the supremum and infimum of f ( x) in ( xi-1 , xi ) for certain partition P of [ a, b] .
n i =1

L ( P, f ) = mi Dxi m Dxi
i =1

(Dxi = xi -1 - xi )

L ( P, f ) m Dxi
i =1

But

Dx
i =1

= ( x1 - x0 ) + ( x2 - x1 ) + ( x3 - x2 ) + .... + ( xn - xn -1 )

Similarity U ( P, f ) M (b - a) m(b - a) L ( P, f ) U ( P, f ) M (b - a)

= xn - x0 = b - a L ( P, f ) m(b - a)

Which shows that the numbers L ( P, f ) and U ( P, f ) form a bounded set. The upper and lower integrals are defined for every bounded function f . Riemann-Stieltjes Integral It is a generalization of the Riemann Integral. Let a ( x) be a monotonically increasing function on [a, b] . a (a ) and a (b) being finite, it follows that a ( x) is bounded on [a, b] . Corresponding to each partition P of [a, b] , we write Da i = a ( xi ) - a ( xi -1 ) ( Difference of values of a at xi & xi -1 ) Q a ( x) is monotonically increasing. \ Da i 0 Let f be a real function which is bounded on [a, b] . Put U ( P, f ,a ) = M i Da i L ( P, f ,a ) = mi Da i
i =1 i =1 n n

Where M i and mi have their usual meanings. Define

f da = inf U ( P, f ,a ) . (i) f da = sup L ( P, f ,a ) . (ii)

a a b

Riemann-Stieltjes Integral

Where the infimum and supremum are taken over all partitions of [a, b] . If

f da = f da ,
a a

we denote their common value by

f da a

or

f ( x) da ( x) . a

This is the Riemann-Stieltjes integral or simply the Stieltjes Integral of f w.r.t. a over [a, b] . If and write f R(a ) . Note The Riemann-integral is a special case of the Riemann-Stieltjes integral when we take a ( x ) = x . Q The integral depends upon f ,a , a and b but not on the variable of integration.

f da exists, we say that f is integrable w.r.t. a , in the Riemann sense,

\ We can omit the variable and prefer to write

In the following discussion f will be assume to be real and bounded, and a monotonically increasing on [a, b] .

f da a

instead of

f ( x) da ( x) . a

Refinement of a Partition Let P and P* be two partitions of an interval [ a, b] such that P P* i.e. every point of P is a point of P* , then P* is said to be a refinement of P . Common Refinement * Let P 1 and P 2 be two partitions of [ a, b] . Then a partition P is said to be their common refinement if P* = P 1P 2. Theorem If P* is a refinement of P , then L ( P, f ,a ) L P* , f ,a .. (i) and U ( P, f ,a )
*

( ) U ( P , f ,a ) . (ii)

Proof Let us suppose that P* contains just one point x * more than P such that xi-1 < x* < xi where xi -1 and xi are two consecutive points of P . Put w1 = inf f ( x) xi-1 x x* xi 1 x* xi w2 = inf f ( x ) It is clear that w1 mi & w2 mi where mi = inf f ( x ) , ( xi -1 x xi ) . Hence * * L P* , f ,a - L ( P, f ,a ) = w1 a ( x ) - a ( xi -1 ) + w2 a ( xi ) - a ( x )

( (x

x xi

-mi [a ( xi ) - a ( xi -1 )]

* * = w1 a ( x ) - a ( xi -1 ) + w2 a ( xi ) - a ( x ) * * - mi a ( xi ) - a ( x ) + a ( x ) - a ( xi -1 ) * * = ( w1 - mi ) a ( x ) - a ( xi-1 ) + ( w2 - mi ) a ( xi ) - a ( x )

Riemann-Stieltjes Integral

Q a is a monotonically increasing function. \ a ( x* ) - a ( xi -1 ) 0 , a ( xi ) - a ( x* ) 0

L P * , f ,a - L ( P, f ,a ) 0 L ( P , f ,a ) L P * , f ,a

which is (i)

If P* contains k points more than P , we repeat this reasoning k times and arrive at (i). Now put W1 = sup f ( x ) ( xi-1 x x* ) and W2 = sup f ( x) ( x* x xi ) Clearly M i W1 & M i W2 Consider U ( P, f ,a ) - U P* , f ,a = M i [a ( xi ) - a ( xi -1 )]

* * - W1 a ( x ) - a ( xi -1 ) - W2 a ( xi ) - a ( x ) * * = Mi a ( xi ) - a ( x ) + a ( x ) - a ( xi -1 ) * * - W1 a ( x ) - a ( xi -1 ) - W2 a ( xi ) - a ( x )

* * = ( M i - W1 ) a ( x ) - a ( xi -1 ) + ( M i - W2 ) a ( xi ) - a ( x ) 0

U ( P , f ,a ) U P * , f ,a

(Q a

is - )

which is (ii)

Theorem Let f be a real valued function defined on [a, b] and a be a monotonically increasing function on [a, b] . Then sup L ( P, f ,a ) inf U ( P, f ,a ) i.e.

f da
a

f da
a

Proof Let P* be the common refinement of two partitions P 1 and P 2 . Then Hence L ( P 1 , f ,a ) U ( P 2 , f ,a ) . (i) If P2 is fixed and the supremum is taken over all P 1 then (i) gives
* L(P 1 , f ,a ) L P , f ,a

U P * , f ,a

U ( P2 , f ,a )

f da
a

U ( P2 , f ,a )

Now take the infimum over all P2

f da
a

f da
a

{{{{{{{{{{{{{{{{{{{{

Riemann-Stieltjes Integral

Theorem (Condition of Integrability or Cauchys Criterion for Integrability.) f R(a ) on [a, b] iff for every e > 0 there exists a partition P such that U ( P, f ,a ) - L ( P, f ,a ) < e Proof Let U ( P, f ,a ) - L ( P, f ,a ) < e . (i) Then L ( P, f ,a )
b

f da f d a
a a

U ( P , f ,a ) U ( P, f ,a ) - f da 0
a b

f da - L ( P, f ,a ) 0
a b b

and

Adding these two results, we have

f da - f da - L ( P, f ,a ) + U ( P, f ,a ) 0
a a

i.e.

f da - f da
a a b b

U ( P, f ,a ) - L ( P, f ,a ) < e <e i.e. for every e > 0 . f R(a )

from (i)

0
b

f da - f da
a a

f da
a

f da
a b b

Conversely, let f R(a ) and let e > 0

b

f da = f da = f da
a a a

Now

f da = inf U ( P, f ,a )
a b

and

f da = sup L ( P, f ,a )
a

There exist partitions P 1 and P 2 such that U ( P2 , f ,a ) - f da <

a b

e .. (ii) 2

U ( P2 , f , a ) - e

< f da

e f da < L ( P1 , f ,a ) + e 2 .. ( iii ) 2 a We choose P to be the common refinement of P 1 and P 2. Then b e U ( P, f ,a ) U ( P2 , f ,a ) < f da + < L ( P 1 , f ,a ) + e L ( P, f ,a ) + e 2 a So that U ( P, f ,a ) - L ( P, f ,a ) < e
and f da - L ( P 1 , f ,a ) <

-----------------------------

Riemann-Stieltjes Integral

Theorem a) If U ( P, f ,a ) - L ( P, f ,a ) < e holds for some P and some e , then it holds (with the same e ) for every refinement of P . b) If U ( P, f ,a ) - L ( P, f ,a ) < e holds for P = { x0 ,...., xn } and si , ti are arbitrary points in [ xi -1, xi ] , then

i =1 n

f ( si ) - f (ti ) Da i < e
b

c) If f R(a ) and the hypotheses of (b) holds, then

f (t )Da - f da
i =1 i i a

<e

Proof a) Let P* be a refinement of P . Then L ( P, f ,a ) L P * , f ,a and U P * , f ,a

L ( P , f ,a ) + U P * , f ,a

U P * , f ,a - L P * , f ,a

Q U ( P, f ,a ) - L ( P, f ,a ) < e
\ U P * , f ,a - L P * , f ,a < e b) P = { x0 ,...., xn } and si , ti are arbitrary points in [ xi -1, xi ] . f ( si ) and f (ti ) both lie in [ mi , M i ] .

) (

) )

U ( P, f ,a )

L P * , f ,a + U ( P , f ,a ) U ( P , f ,a ) - L ( P , f ,a )

) (

f ( si ) - f (ti ) M i - mi f ( si ) - f (ti ) Da i M i Da i - mi Da i

xi 1 si
n

ti

xi

i =1 i =1 n

f ( si ) - f (ti ) Da i

M
i =1

Da i - mi Da i
i =1

f ( si ) - f (ti ) Da i U ( P, f ,a ) - L ( P, f ,a )

Q U ( P , f ,a ) - L ( P , f ,a ) < e
\ c)

i =1

f (si ) - f (ti ) Da i < e

Q mi f (ti ) M i \

m Da
i

f (t ) Da M Da L ( P, f ,a ) f (t ) Da U ( P, f ,a )
i

and also L ( P, f ,a ) Using (b), we have

f da a
b

U ( P, f ,a )

f (t ) Da - f da
i i a

< e

-----------------------------

Riemann-Stieltjes Integral

Theorem If f is continuous on [a, b] then f R(a ) on [a, b] . Proof Let e > 0 be given. Choose b > 0 so that [a (b) - a (a)] b < e f is continuous on [a, b] f is uniformly continuous on [a, b] . There exists a d > 0 such that f ( s) - f (t ) < b if x [a, b] , t [a, b] and x - t < d ..(i) If P is any partition of [a, b] such that Dxi < d for all i , (i = 1,2,...., n) then (i) implies that M i - mi b U ( P, f ,a ) - L ( P, f ,a ) = M i Da i - mi Da i = ( M i - mi ) Da i

b Da i = b [a (b) - a (a )] < e f R (a ) by Cauchy Criterion.

Theorem If f is monotonic on [a, b] , and if a is continuous on [a, b] , then f R(a ) . ( Monotonicity of a still assumed. ) Proof Let e > 0 be a given positive number. For any positive integer n , choose a partition P = { x0 , x1,....., xn } of [a, b] such that a (b) - a (a ) Da i = , i = 1,2,...., n n This is possible because a is continuous and monotonic increasing on the closed interval [a, b] and thus assumes every value between its bounds, a (a ) and a (b) . Let f be monotonic increasing on [a, b] , so that its lower and upper bounds mi , M i in [ xi -1, xi ] are given by mi = f ( xi -1 ) , M i = f ( xi ) , i = 1,2,...., n \ U ( P, f ,a ) - L ( P, f ,a ) = ( M i - mi ) Da i
i =1 n

a (b) - a (a ) n = [ f ( xi ) - f ( xi-1 )] n i =1 a (b) - a (a ) = [ f (b) - f (a)] n <e if n is taken large enough.

f R(a ) on [a, b] . Note: f R(a ) when either i) f is continuous and a is monotonic, or ii) f is monotonic and a is continuous, of course a is still monotonic.

Riemann-Stieltjes Integral

Properties of Integral i) If f R(a ) on [a, b] , then cf R(a ) for every constant c and cf da = c f da . a a Proof Q f R (a ) \ $ a partition P such that U ( P, f ,a ) - L ( P, f ,a ) < e
n n b b

, where e is an arbitrary +ive number.

Now

U ( P, cf ,a ) = cM i Da i = c M i Da i
i =1 n i =1 n

& L ( P, cf ,a ) = cmi Da i = c mi Da i U ( P , f ,a ) - L ( P , f ,a ) = c M i Da i - mi Da i
i =1 i =1

Q U ( P, cf ,a ) = c U ( P, f ,a )

cf R(a )

= c U ( P, f ,a ) - L ( P, f ,a ) < ce = e1 & L ( P, cf ,a ) = c L ( P , f ,a )

\ inf U ( P, cf ,a ) = c inf U ( P, f ,a ) & sup L ( P, cf ,a ) = c sup L ( P, f ,a ) where infimum and supremum are taken over all P on [a, b] .

cf da = c f da
a a

& and
b

cf da = c f da
a b b a

cf da = cf da
a a b

f da = f d a
a a

cf da = c f da a a

ii) If f1 R (a ) and f 2 R (a ) on [a, b] , then f1 + f 2 R (a ) and

( f1 + f 2 ) da = f1 da + f 2 da . a a a
Proof If f = f1 + f 2 and P is any partition of [a, b] , we have mi + mi mi M i M i + M i where M i , mi , M i, mi and M i , mi are the bounds of f1 , f 2 and f respectively in [ xi-1, xi ] . Multiplying throughout by Da i and adding the inequalities for i = 1,2,...., n , we get L ( P , f1 ,a ) + L ( P , f 2 ,a ) L ( P , f ,a ) U ( P, f ,a ) U ( P, f1 ,a ) + U ( P , f 2 ,a ) .....(i) Since f1 R (a ) and f 2 R (a ) on [a, b] therefore $ e > 0 and there are partitions P 1 and P 2 such that .. (ii) and U ( P2 , f 2 ,a ) - L ( P2 , f 2 ,a ) < e These inequalities hold if P 1 and P 2 are replaced by their common refinement P . U (P 1 , f1 ,a ) - L ( P 1 , f1 ,a ) < e

Riemann-Stieltjes Integral

(ii) U ( P, f1 ,a ) + U ( P, f 2 ,a ) - L ( P, f1,a ) + L ( P, f 2 ,a ) < 2e Using (i) we have U ( P, f ,a ) - L ( P, f ,a ) < 2e which proves that f R(a ) on [a, b] With the same partition P , we have U ( P, f1,a ) < and
b

f1 da + e a f 2 da + e a f1 da + f 2 da + 2e a a
b b b b

U ( P, f 2 ,a ) <

Hence (i) implies that f da a U ( P, f ,a ) <

Q e is arbitrary, we conclude that

f da a f da a
b b b b

f1 da + f 2 da a a f1 da + f 2 da a a
b b b b

Similarly if we consider the lower sums we arrive at

Combining the above two results, we have f da = f1 da + f 2 da a a a iii) If f1 ( x) f 2 ( x) on [a, b] , then f1 da a Proof Let f ( x ) 0 , then M i 0 and \ f da a
b b b

f 2 da a

U ( P , f ,a ) 0

0 \ f 2 - f1 0 0 f 2 da - f1 da a a
b b

Q f1 f 2 Note (i)

( f 2 - f1 ) da a
f1 da a
b b

f 2 da a

(f (f

+ g ) ( x ) = f ( x ) + g ( x ) sup f + sup g sup ( f + g ) sup f + sup g + g ) ( x) = f ( x) + g ( x) inf f + inf g inf ( f + g ) inf f + inf g

(ii)

10

Riemann-Stieltjes Integral

iv) If f R(a ) on [a, b] and if a < c < b , then f R(a ) on [a, c ] and on [c, b] and f da = f da + f da a a c Proof Since f R(a ) on [a, b] , therefore for e > 0 , $ a partition P such that U ( P, f ,a ) - L ( P, f ,a ) < e Let P* be the refinement of P such that P* = P {c}
* * b c b

\ L ( P, f ,a ) L P* , f ,a U P* , f ,a U ( P, f ,a ) ....... (i) .(ii)

* Let P 1,P 2 denote the sets of points of P between [ a, c ] , [c , b] respectively.

( ) ( ) U ( P , f ,a ) - L ( P , f ,a ) U ( P , f ,a ) - L ( P, f ,a ) < e

Clearly P P* = P 1,P 2 are partitions of [ a, c ] , [c , b] respectively and 1P 2. Also and U P* , f ,a = U ( P 1 , f ,a ) + U ( P 2 , f ,a ) .. (iii)

* 1 2

\ {U ( P 1 , f ,a ) - L ( P 1 , f ,a )} + {U ( P 2 , f ,a ) - L ( P 2 , f ,a )}

( ) L ( P , f ,a ) = L ( P , f ,a ) + L ( P , f ,a ) (iv) (

= U P * , f ,a - L P * , f ,a < e Since each bracket on the left is non-negative, it follows that U (P 1 , f ,a ) - L ( P 1 , f ,a ) < e

) (

and U ( P2 , f ,a ) - L ( P2 , f ,a ) < e f R (a ) on [a, c ] and on [c, b] . We know that for any functions f1 and f 2 , if f = f1 + f 2 , then inf f inf f1 + inf f 2 and sup f sup f1 + sup f 2
* Now for any partitions P 1,P 2 of [ a, c ] , [c , b] respectively, if P = P 1P 2 , then

U P* , f ,a = U ( P 1 , f ,a ) + U ( P 2 , f ,a )

Hence on taking the infimum for all partitions, we get

f da
a b

f da + f da
a c

But since f R(a ) on [a, c ] , [c, b] , [a, b ] \ Again f da a

c

L P* , f ,a = L ( P 1 , f ,a ) + L ( P 2 , f ,a )

f da + f da a c

. (v)

and on taking the supremum for all partitions, we get

f da
a

f da + f da
a c

But since f R(a ) on [a, c ] , [c, b] , [a, b ] \ (v) and (vi) imply that f da a
b

f da a =

f da + f da a c
b

(vi)

f da + f da a c

Riemann-Stieltjes Integral

11

v) If

f R(a ) on [a, b] and

f ( x ) M on [a, b] , then
M [a (b) - a (a )]

f da
a

Proof We know that f da a But

b

U ( P, f ,a ) = M i Da i M Da i = a (b) - a (a ) M [a (b) - a (a ) ]

Da

f da
a

vi) If f R (a1 ) and f R(a 2 ) , then f R(a1 + a 2 ) and f d (a1 + a 2 ) = f da1 + f da 2 a a a f d (ca ) = c f da a a Proof Since f R (a1 ) and f R(a 2 ) , therefore for e > 0 , there exists partitions P 1, P 2 of [ a, b] such that e U (P 1 , f ,a1 ) - L ( P 1 , f ,a1 ) < 2 e and U ( P2 , f ,a 2 ) - L ( P2 , f ,a 2 ) < 2 Let P = P 1P 2 e \ U ( P, f ,a1 ) - L ( P, f ,a1 ) < 2 .. (i) e & U ( P, f ,a 2 ) - L ( P, f ,a 2 ) < 2 Let mi , M i be bounds of f in [ xi -1, xi ] Take a = a1 + a 2 Da i = Da1i + Da 2 i \ U ( P , f ,a ) = M i Da i = M i Da1i + Da 2 i
b b b b b

and if f R(a ) and c is a positive constant, then f R(ca ) and

= U ( P, f ,a1 ) + U ( P, f ,a 2 ) L ( P, f ,a ) = L ( P, f ,a1 ) + L ( P, f ,a 2 ) \ U ( P, f ,a ) - L ( P, f ,a ) = U ( P, f ,a1 ) - L ( P, f ,a1 ) + U ( P, f ,a 2 ) - L ( P, f ,a 2 ) e e < + = e by (i) 2 2 f R (a ) where a = a1 + a 2

Similarly

12

Riemann-Stieltjes Integral

To prove the second part, we notice that f da a

b

= inf U ( P, f ,a ) = inf {U ( P, f ,a1 ) + U ( P, f ,a 2 )} = f da1 + f da 2 a a f da1 + f da 2 a a f da1 + f da 2 a a

b b b b b b b b

inf U ( P, f ,a1 ) + inf U ( P, f ,a 2 )

.... (ii)

Similarly by taking the supremum of lower sum of partition we arrive that f da a From (ii) and (iii) f da a
b b b

..... (iii)

i.e.

Now Q f R (a ) \ for e > 0 , $ a partition P of [a, b] such that U ( P, f ,a ) - L ( P, f ,a ) < e . (iv) Let a = ca then Da i = D (ca i ) = cDa i U ( P, f ,a ) = M i Da i = M i ( cD a i ) = c M i Da i

f d (a1 + a 2 ) = f da1 + f da 2 a a a

Q a = a1 + a 2

= c U ( P , f ,a ) Similarly, L ( P, f ,a ) = c L ( P, f ,a ) U ( P, f ,a ) - L ( P, f ,a ) = c {U ( P, f ,a ) - L ( P, f ,a )} < c e f R (a ) where a = ca

Also f da = inf U ( P, f ,a ) a
b

by (iv)

= inf c U ( P, f ,a ) = c inf U ( P, f ,a )
= c f da
a b

and f da = sup L ( P, f ,a ) a
b

= sup cU ( P, f ,a ) = c supU ( P, f ,a )
= c f da
a b

Hence f da = c f da a a
b b

where a = ca

Riemann-Stieltjes Integral

13

Lemma If M & m are the supremum and infimum of f and M , m are the supremum & infimum of f on [a, b] then M - m M - m . Proof Let x1, x2 [a, b] , then f ( x1 ) - f ( x2 ) f ( x1 ) - f ( x2 ) (A) Q M and m denote the supremum and infimum of f ( x) on [a, b] \ f ( x) M & f ( x ) m " x [ a, b] Q x1 , x2 [a, b] \ f ( x1 ) M and f ( x2 ) m f ( x1 ) M and - f ( x2 ) - m f ( x1 ) - f ( x2 ) M - m ... (i) Interchanging x1 & x2 , we get - [ f ( x1 ) - f ( x2 )] M - m .. (ii) (i) & (ii)

f ( x1 ) - f ( x2 ) M - m
f ( x1 ) - f ( x2 ) M - m by eq. (A) ...(I)

Q M and m denote the supremum and infimum of f ( x ) on [a, b] \ f ( x) M and f ( x) m " x [a, b] $ e > 0 such that f ( x1 ) > M - e .. (iii) and f ( x2 ) < m + e - f ( x2 ) + e > - m .. (iv) From (iii) and (iv), we get f ( x1 ) - f ( x2 ) + e > M - m - e 2e + f ( x1 ) - f ( x2 ) > M - m

Q e is arbitrary \ M - m f ( x1 ) - f ( x2 ) .... (v) Interchanging x1 & x2 , we get M - m - ( f ( x1 ) - f ( x2 ) ) (vi)

Combining (v) and (vi), we get M - m f ( x1 ) - f ( x2 ) . (II) From (I) and (II), we have the require result M - m M - m Theorem If f R(a ) on [a, b] , then

f R (a ) on [a, b] and

f da
a

f da .

Proof Q f R (a ) \ given e > 0 $ a partition P of [a, b] such that U ( P, f ,a ) - L ( P, f ,a ) < e i.e.

Where M i and mi are supremum and infimum of f on [ xi -1, xi ] Now if M i and mi are supremum and infimum of f on [ xi -1, xi ] then M i - mi M i - mi

Da i - mi Da i = ( M i - mi ) Da i < e

14
i i i

Riemann-Stieltjes Integral

( M - m ) Da ( M - m ) Da U ( P , f ,a ) - L ( P, f ,a ) U ( P , f ,a ) - L ( P , f ,a ) < e
i i i

Take c = +1 or -1 to make c f da 0 Then Also

f R (a ) .

f da
a

= c f da . (i)
a

c f ( x)
c f da a
b

f ( x)

" x [ a, b]
b

f da c f da
a

f da . (ii)

From (i) and (ii), we have

f da
a

f da

Theorem If f R(a ) on [a, b] , then f 2 R (a ) on [a, b] . Proof Q f R(a ) f R (a ) f ( x) < M " x [ a, b] Q f R (a ) \ given e > 0 , $ a partition P of [a, b] such that U ( P, f ,a ) - L ( P, f ,a ) < e (i) 2M If M i & mi denote the sup. & inf. of f on [ xi -1, xi ] then M i2 & mi2 are the U P, f 2 ,a - L P, f 2 ,a = M i2 - mi2 Da i " x [ a, b]

sup. & inf. of f 2 on [ xi -1, xi ] .

) (

Q f ( x) f ( x) M
and f 2 = f \ M i M & mi M
2

= ( M i + mi )( M i - mi ) Da i

U P, f 2 ,a - L P, f 2 ,a ( M + M ) ( M i - mi ) Da i = 2 M ( M i - mi ) Da i = 2M U ( P, f ,a ) - L ( P, f ,a ) < 2M

) (

f 2 R(a )
Corollary If f R(a ) & g R (a ) on [a, b] then fg R(a ) on [a, b] . Proof Q f R (a ) , g R (a ) \ f + g R(a ) , f - g R (a )

e =e 2M

( f + g )2 R(a ) , ( f - g ) 2 R (a ) ( f + g )2 - ( f - g ) 2 R (a ) 4 fg R (a ) and ultimately fg R(a ) on [a, b]

Riemann-Stieltjes Integral

15

Theorem Assume a increases monotonically and a R on [a, b] . Let f be bounded real function on [a, b] . Then f R(a ) iff f a R . In that case f da = f ( x) a ( x ) dx a a Proof Q a R on [a, b] \ given e > 0 $ a partition P of [a, b] such that U ( P,a ) - L ( P,a ) < e .... (i) The Mean-value theorem furnishes point ti [ xi -1, xi ] such that Da i = a ( xi ) - a ( xi -1 ) = a (ti ) Dxi for i = 1,2,...., n . (ii) If si [ xi -1, xi ] , then form (i) we have
b b

| Previously proved at page 6 a ( si ) Dxi - a (ti ) Dxi < e a (si ) - a (ti ) Dxi < e (iii)

Put M = sup f ( x) and consider = =

Thus U ( P, f ,a ) - U ( P, f a ) M e ... (v) Q (i) remains true if P is replaced by any refinement \ (v) also remains true
b

.... (iv) by (iii) f ( si ) Da i f ( si )a ( si ) Dxi + M e for all choices of si [ xi -1, xi ] U ( P, f ,a ) U ( P, f a ) + M e The same arguments leads from (A) to U ( P, f a ) U ( P, f ,a ) + M e

Me

f ( si ) Dai - f ( si )a ( si ) Dxi .. (A) f ( si )a (ti ) Dxi - f ( si )a ( si ) Dxi by (ii) f ( si ) (a (ti ) - a ( si ) ) Dxi M (a (ti ) - a ( si ) ) Dxi

f da - f ( x)a ( x) dx
a a b

Me

Q e was arbitrary
\ f da a

f ( x)a ( x) dx a

for any bounded f .

Using the same argument, we can prove from (iv ) by considering the infimum of f ( x) that

f da
a

f ( x)a ( x) dx
a

Hence

f da = f da
a a

f ( x)a ( x) dx = f ( x)a ( x) dx
f a R(a ) .
a a

Equivalently

f R(a )

16

Riemann-Stieltjes Integral

Theorem (Change of Variable) Suppose j is a strictly increasing continuous function that maps an interval [ A, B] onto [a, b] . Suppose a is monotonically increasing on [a, b] and f R(a ) on [a, b] . Define b and g on [ A, B ] by b ( y ) = a ( j ( y ) ) , g ( y ) = f (j ( y ) ) then g R ( b ) and Proof
b a f
a(f(y))= b(y)

g d b = f da . A a

y [A,B]

x=f(y)

f [a,b]

f(f(y))= g(y)

R g

Q The value taken by f on [ xi -1, xi ] are exactly the same as those taken by g on [ yi -1 , yi ] , we see that U ( Q, g , b ) = U ( P, f ,a ) and L ( Q, g , b ) = L ( P, f ,a ) Q f R(a ) on [a, b] \ given e > 0 , we have U ( P, f ,a ) - L ( P, f ,a ) < e U ( Q, g , b ) - L ( Q, g , b ) < e
g R( b ) and g d b = f da A a
B b

To each partition P = { x0 ,....., xn } of [a, b] corresponds a partition Q = { y0 ,....., yn } of [ A, B ] because j maps [ A, B ] onto [a, b] . xi = j ( yi ) All partitions of [ A, B ] are obtained in this way.

{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{{

Riemann-Stieltjes Integral

17

INTEGRATION AND DIFFERENTIATION Theorem (Ist Fundamental Theorem of Calculus) Let f R on [a, b] . For a x b , put F ( x) = f (t ) dt , then F is continuous
a x

on [a, b] ; furthermore, if f is continuous at point x0 of [a, b] , then F is differentiable at x0 , and F ( x0 ) = f ( x0 ) . Proof Q f R \ f is bounded. Let f (t ) M for t [a, b] If a x < y b , then F ( y) - F ( x) = = =

a
x

f (t ) dt - f (t ) dt
a a

f (t ) dt + f (t ) dt - f (t ) dt
a x a

f (t ) dt
x

f (t ) dt M dt = M ( y - x)
x

for e > 0 provided M y - x < e e i.e. F ( y ) - F ( x ) < e whenever y - x < M This proves the continuity (and, in fact, uniform continuity) of F on [a, b] . Next, we have to prove that if f is continuous at x0 [a , b] then F is differentiable at x0 and F ( x0 ) = f ( x0 ) F (t ) - F ( x0 ) i.e. lim = f ( x0 ) t x0 t - x0 Suppose f is continuous at x0 . Given e > 0 , $ d > 0 such that f (t ) - f ( x0 ) < e if t - x0 < d where t [a, b] f ( x0 ) - e < f (t ) < f ( x0 ) + e if x0 - d < t < x0 + d

F ( y ) - F ( x) < e

( f ( x ) - e ) dt
0 x0

< <

f (t ) dt
x0

< <

( f ( x ) + e ) dt
0 x0

t x0+d

a x0 d

x0

( f ( x0 ) - e ) dt
x0

f (t ) dt
x0

( f ( x0 ) + e ) dt
x0

< F (t ) - F ( x0 ) < ( f ( x0 ) + e ) (t - x0 ) F (t ) - F ( x0 ) f ( x0 ) - e < < f ( x0 ) + e t - x0

F (t ) - F ( x0 ) - f ( x0 ) < e t - x0 F (t ) - F ( x0 ) lim = f ( x0 ) t x0 t - x0 F ( x0 ) = f ( x0 ) ---------------------------------

( f ( x0 ) - e ) (t - x0 )

18

Riemann-Stieltjes Integral

Theorem (IInd Fundamental Theorem of Calculus) If f R on [a, b] and if there is a differentiable function F on [a, b] such that F = f , then f ( x) dx = F (b ) - F (a ) a Proof Q f R on [a, b] \ given e > 0 , $ a partition P of [a, b] such that U ( P, f ) - L ( P, f ) < e Q F is differentiable on [a, b] \ $ ti [ xi-1 , xi ] such that F ( xi ) - F ( xi -1 ) = F (ti )Dxi F ( xi ) - F ( xi -1 ) = f (ti ) Dxi for i = 1,2,...., n
b

Q F = f

f (t ) Dx
i =1 i

= F (b) - F (a )
b

Q if f R (a ) then

F (b) - F (a ) - f ( x) dx < e

f (t ) Da -
i i

b a

f da < e

Q e is arbitrary
\ f ( x ) dx = F (b ) - F (a ) a
b

Theorem (Integration by Parts) Suppose F and G are differentiable function on [a, b] , F = f R and G = g R then F ( x ) g ( x) dx = F (b) G (b) - F (a ) G (a ) - f ( x ) G ( x) dx a a Proof Put H ( x) = F ( x) G ( x ) H = F ( x) G ( x) + F ( x ) G( x) = h Now Q H R and h R on [a, b] \ By applying the fundamental theorem of calculus to H and its derivative h , we have h dx = H (b) - H (a ) a
b b b

[ F ( x) G ( x) + F ( x) G( x)] dx = H (b) - H (a) a

f ( x ) G ( x) dx + F ( x ) g ( x ) dx = F (b) G (b) - F (a ) G (a ) a a F ( x) g ( x ) dx = F (b) G (b ) - F (a ) G (a ) - f ( x) G ( x ) dx a a ----------------------------------
b b b b

Made by: Atiq ur Rehman (atiq@mathcity.tk) Available online at http://www.mathcity.tk in PDF Format. Page Setup: Legal ( 8 14 ) Printed: 15 April 2004 (Revised: June 08, 2004.)
1 2

Riemann-Stieltjes Integral

19

Question Show that the function f defined on [ 0,1] by 1 f ( x) = 0 ; x is rational ; x is irrational

is not integrable on [ 0,1] Solution For any partition P of [ 0,1] , mk = 0 , M k = 1 S ( P, f ) = M k Dxk = Dxk = 1 - 0 = 1 and so that i.e. L ( P, f ) = mk Dxk = 0
k =1 k =1 n k =1 n n

f dx = 1
0 1 1 0 0

f dx = 0
0

f dx f dx

f is not integrable on [ 0,1] .

Question p Show that f ( x ) = sin x is Riemann integrable over 0, . 2 Solution np p p 3p p Take P = 0, , , ,....., by dividing 0, into n equal parts. 2n 2n n 2n 2 kp (k - 1)p Then M k = sin , mk = sin 2n 2n kp (k - 1)p p S ( P, f ) - L ( P, f ) = sin - sin 2n 2n 2n p p <e for n > n0 = 2n 2e p f is Riemann integrable over 0, . 2 Question 1 Show that f ( x) = x 0 is integrable on [ 0,1] . ; x is rational , 0 < x 1 ; x is irrational

Solution f is continuous at each irrational. And rational numbers are dense in [ 0,1] . Also L ( P, f ) = 0 for any partition P of [ 0,1] so that Q f 0

f dx = 0
0

\ S ( P, f ) 0

f da
0

0 ... (i)

20

Riemann-Stieltjes Integral

Q There are only finite number of points \ Suppose f ( x )

e for k values of x in [ 0,1] 2 e Take P P . 1 such that 1 < 2k

Consider S ( P 1 , f ) = M i ( xi - xi -1 )
i =1 n

p q e p (rationals) for which f = q q p 2

There are at most k values for which

S(P 1, f ) =

k values

e e M i 1 . For all other values M i > . 2 2 M i ( xi - xi -1 ) + M i ( xi - xi -1 )

other values

Q e is arbitrary \ S (P 1, f ) 0
and

e e e e k + ( xi - xi -1 ) < + = e 2k 2 2 2

f dx 0 (ii)
0

By (i) and (ii), we have

f dx = 0
0

Hence

f dx = 0 0

Note If f is integrable then f is also integrable but the converse is false. For example, let f be a function defined on [a, b] by 1 ; x [ a, b] f ( x) = -1 ; otherwise Then f is Riemann-integrable but f is not.

References:

(1) Lectures (Year 2003-04) Prof. Syyed Gul Shah (2) Book Mathematical Analysis

Chairman, Department of Mathematics. University of Sargodha, Sargodha.

Tom M. Apostol (John Wiley & Sons, Inc.)

Made by: Atiq ur Rehman (atiq@mathcity.org) Available online at http://www.mathcity.org in PDF Format. Page Setup: Legal ( 8 1 2 14 ) Printed: 15 April 2004 (Revised: March 19, 2006.) Submit error or mistake at http://www.mathcity.org/error