MA105: Calculus

Lecture 8 (D1)

Shripad M. Garge
IIT Bombay, Mumbai.

August 20, 2018.

Shripad M. Garge, IITB MA105: D1-08

Recall

Riemann integration,

examples,

integrability theorems,

August 22.

Shripad M. Garge, IITB MA105: D1-08

Integration: Examples

Let us see one more example to see the method clearly.

We take f (x) = x on a closed and bounded interval [a, b].

Let Ṗ be a tagged partition of the interval [a, b] given by the

subintervals [xi , xi+1 ] and tags ti ∈ [xi , xi+1 ] for i = 0, 1, . . . , n − 1.
We observed that
n−1
X n−1
X
`= xi (xi+1 − xi ) ≤ S(f , Ṗ) ≤ xi+1 (xi+1 − xi ) = u
i=0 i=0

and also that

n−1
b 2 − a2 1X
S(f , Ṗ) − ≤ (xi+1 − xi )2 .
2 2
i=0

Shripad M. Garge, IITB MA105: D1-08

Integration: Examples

Observe that if kṖk < δ then (xi+1 − xi ) < δ for every i and hence
n−1
X n−1
X
(xi+1 − xi )2 < δ(xi+1 − xi ) = δ(b − a).
i=0 i=0

2
Now, for an  > 0 we choose δ < b−a so that kṖk < δ implies that

b 2 − a2
S(f , Ṗ) − < .
2

Thus the identity function is integrable and its integral is the

number that we expect it to be, (b 2 − a2 )/2.

Note that it was useful to take tags to be the global extrema in

each of the subintervals [xi , xi+1 ]. We will visit this theme again.

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

Theorem
Z b Z b
kf (x)dx = k f (x)dx.
a a
Z b Z b Z b
(f ± g )(x)dx = f (x)dx ± g (x)dx.
a a a

If f (x) ≤ g (x) for all x ∈ [a, b] then

Z b Z b
f (x)dx ≤ g (x)dx.
a a

A step function is always integrable and its integral is given by

summing the integrals of its steps.

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

Theorem
If f ∈ R[a, b] then f is bounded on [a, b].

Proof.
Rb
Let f : [a, b] → R be unbounded with a f (x)dx = L. There exists
δ > 0 such that for any tagged partition Ṗ with kṖk < δ,

|S(f , Ṗ) − L| < 1

and hence |S(f , Ṗ)| < |L| + 1.

Let P = (x0 < x1 < · · · < xn ). We will choose suitable tags so as

to get a contradiction (because we want to prove that f is
bounded!).

(Proof continued on the next page)

Shripad M. Garge, IITB MA105: D1-08
Integrability theorems

Proof. (continued)
Since f is unbounded on [a, b], it is unbounded on at least one
subinterval, say on [xk , xk+1 ].
Choose any tags ti ∈ [xi , xi+1 ] for i 6= k and then choose
tk ∈ [xk , xk+1 ] such that
X
|f (tk )(xk+1 − xk )| > |L| + 1 + f (ti )(xi+1 − xi ) .
i6=k

Then by triangle inequality (|A + B| ≥ |A| − |B|), we get

X
|S(f , Ṗ)| ≥ |f (tk )(xk+1 − xk )| − f (ti )(xi+1 − xi ) > |L| + 1.
i6=k

Contradiction! 

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

Theorem
(Squeeze theorem) A function f : [a, b] → R belongs to
R[a, b] if and only if for every  > 0 there are functions
α, β ∈ R[a, b] such that α(x) ≤ f (x) ≤ β(x) for all x ∈ [a, b]
and Z b
(β − α) dx < .
a

If f : [a, b] → R is continuous then f ∈ R[a, b].

We will use the squeeze theorem and prove that every continuous
f : [a, b] → R is integrable.

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

Proof.
Let f : [a, b] → R be continuous. For any  > 0, let
1 = /(b − a). For this 1 , we get a δ > 0 such that

|x − y | < δ =⇒ |f (x) − f (y )| < 1 .

Let P = (x0 < x1 < · · · < xn ) be a partition of [a, b] with kPk < δ.
For each i, let ui be the minimum value of f on [xi , xi+1 ] and let vi
be the maximum value, that is,

ui ≤ f (x) ≤ vi for every x ∈ [xi , xi+1 ].

Define two step functions α, β : [a, b] → R by declaring that on

each [xi , xi+1 ], α = ui and β = vi .

(Proof continued on the next page)

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

Proof. (continued)
Then α, β ∈ R[a, b] and α(x) ≤ f (x) ≤ β(x). Moreover
Z b X
(β − α) = (vi − ui )(xi+1 − xi ) < 1 (b − a) = .
a i

Using squeeze theorem, we get that f ∈ R[a, b]. 

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

Theorem
If f : [a, b] → R is bounded and has only finitely many (or
even countably many) discontinuities then f ∈ R[a, b].

If c ∈ (a, b) then
Z b Z c Z b
f (x)dx = f (x)dx + f (x)dx.
a a c

The last result can be generalised to finitely many points

c1 < c2 < · · · < cr in (a, b).

Note that this last result can be proved from the scratch, using
only the definitions. We may have used it before stating it, but it
is ok as we are not going to use any of the previous results in
proving this result.
Shripad M. Garge, IITB MA105: D1-08
Fundamental theorem of calculus (First form)

The fundamental theorem of calculus combines differentation and

integration.

Theorem
Suppose that there is a finite subset E of [a, b] and functions
f , F : [a, b] → R satisfying
1 F is continuous on [a, b],
2 F 0 (x) = f (x) for all x ∈ [a, b] − E ,
3 f belongs to R[a, b].
Then we have Z b
f (x)dx = F (b) − F (a).
a

Shripad M. Garge, IITB MA105: D1-08

Fundamental theorem of calculus (First form)

Proof.
Steps:
1 We prove the result when E = {a, b}.
2 For any  > 0 we get δ > 0 such that for any tagged partition
Ṗ with norm less than δ,
Z b
|S(f , Ṗ) − f | < .
a

3 Using MVT for F over the subintervals given by P, we can

select tags such that

S(f , Ṗ) = F (b) − F (a).

Rb
4
a f (x) dx = F (b) − F (a).

Shripad M. Garge, IITB MA105: D1-08

Integral computations

Now, we can compute several integrals.

f : [a, b] → R, f (x) = x, F (x) = 12 x 2 ,
Z b
1
f (x)dx = (b 2 − a2 ).
a 2

f : [a, b] → R, f (x) = 1/(x 2 + 1), F (x) = arctan x.

Z b
f (x)dx = arctan(b) − arctan(a).
a

f : [−a, a] → R, f (x) = x/|x| if x 6= 0, and f (0) = 0. Then

F (x) = |x| has the property that F 0 = f on [−a, 0) ∪ (0, a]
and hence Z a
f (x)dx = |a| − |(−a)| = 0.
−a

Shripad M. Garge, IITB MA105: D1-08

Fundamental theorem of calculus (Second form)

In the first form of the FTC, we computed

Z b
F 0 (x)dx
a

in terms of values of F . Thus, we integrated the differential of F .

Now, in the second form, we are going to differentiate an integral.

If f ∈ R[a, b], we define the indefinite integral of f to be the

function Z z
F (z) = f (x) dx.
a
We will see that when this function is differentiable, its derivative
is given in terms of values of f . But, first, we prove that it is
continuous.

Shripad M. Garge, IITB MA105: D1-08

Fundamental theorem of calculus (Second form)

Theorem
The indefinite integral F is continuous on [a, b].
In fact, there exists M > 0 such that for all z, w ∈ [a, b]

|F (z) − F (w )| < M|z − w |.

Proof.
Rz
If w , z ∈ [a, b] and w ≤ z then F (z) − F (w ) = w f (x) dx.
Further, since f is bounded, there exists an M > 0 such that
Z z
−M(z − w ) < f (x) dx < M(z − w ).
w

Hence
|F (z) − F (w )| < M|z − w |.

Shripad M. Garge, IITB MA105: D1-08

Fundamental theorem of calculus (Second form)

Theorem
Let f ∈ R[a, b] and let f be continuous at c ∈ (a, b). Then the
indefinite integral F is differentiable at c and F 0 (c) = f (c).

Proof.
For  > 0, there exists δ > 0 such that |x − c| < δ =⇒
f (c) −  < f (x) < f (c) + . Then for 0 < h < δ
c+h
F (c + h) − F (c)
Z
1
− f (c) = f (x) dx − f (c) < .
h h c

Thus, limh→0+ F (c+h)−F

h
(c)
exists and is equal to f (c). The left
hand limit is computed in the same way and we get the result.

Shripad M. Garge, IITB MA105: D1-08

FTC in a nutshell

Z b  
d
F (t) dt = F (b) − F (a)
a dx

and

Z x 
d
f (t) dt = f (x).
dx a

Shripad M. Garge, IITB MA105: D1-08

Limitations of FTC

If a function f : [a, b] → R is Riemann integrable but is not F 0 for

any F : [a, b] → R, then Fundamental theorem of calculus cannot
help us in computing its integral.

Theorem
If F : [a, b] → R is differentiable then F 0 satisfies the intermediate
value property.

Thomae’s function f : [0, ∞) → R is defined by f (x) = 0 if x is

irrational and f (m/n) = 1/n where m, n ∈ N and m, n have no
common factor.

This function is continuous at all irrationals, so f is Riemann

integrable on any [a, b] ⊆ [0, ∞). But, by above theorem, it is not
F 0 for any F .

Shripad M. Garge, IITB MA105: D1-08

Thomae’s function

Shripad M. Garge, IITB MA105: D1-08

Integration: lower and upper sums

Let f : [a, b] → R be bounded and let P = (x0 < · · · < xn ) be a

partition of
 [a, b]. On each subinterval [xi ,xi+1 ], let
ui = min f (x) : x ∈ [xi , xi+1 ] , vi = max f (x) : x ∈ [xi , xi+1 ] .
We define lower P and upper sums of f with respect Pto P to be the
sums L(f , P) = i ui (xi+1 − xi ) and U(f , P) = i vi (xi+1 − xi ).
Theorem
A bounded f : [a, b] → R is Riemann integrable if and only if

sup{L(f , P)} = inf {U(f , P)}.

P P

Rb
In this case, a f (x) dx is the supremum of the lower sums which
is the same as the infimum of the upper sums.

Shripad M. Garge, IITB MA105: D1-08

Integration: lower and upper sums

These sums help us in computing integrals when we cannot use the

fundamental theorem of calculus.

It is customary to consider partitions Pn consisting of subintervals

of equal length (b − a)/n and take the limits of the lower and
upper sums.

Now, these are real limits, unlike the earlier Riemann sums, and
our expertise in limits of sequences should help us here!

Let us do some calculations.

Shripad M. Garge, IITB MA105: D1-08

Integration: computations

Consider the function f : [a, b] → R, f (x) = x. We take

Pn = (a < · · · < a + i(b−a)
n < · · · < b) for 0 ≤ i ≤ n. Then

L(f , Pn ) =? and U(f , Pn ) =?

One sees that

1
lim L(f , Pn ) = lim U(f , Pn ) = (b 2 − a2 ).
n→∞ n→∞ 2
If you look at L(f , Pn ) or U(f , Pn ) carefully, you will realise that
while computing their limits, you have actually verified
convergence of a series.

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

In a reverse sort of a way, one can compute sum of a series by

identifying it as limit of certain Riemann sums.

For example, fix an r > 0 and consider

1r + 2r + · · · + nr
Sn = .
nr +1
What is limn→∞ Sn ?

r
Observe that Sn = i ni n1 , and this is the Riemann sum for our
P
canonical partition for the function x r over [0, 1]. We get that
Z 1
1
lim Sn = x r dx = .
n→∞ 0 r +1

Shripad M. Garge, IITB MA105: D1-08

Integrability theorems

(Integration by parts)
Z b Z b
0
FG = F (b)G (b) − F (a)G (a) − F 0G .
a a

(Integration by substitution) If f : [a, b] → R is continuous

and φ : [α, β] → [a, b] is continuously differentiable then
Z φ(β) Z β
f (x) dx = f (φ(t))φ0 (t) dt.
φ(α) α

(Leibniz’s rule for integrals) If f : [a, b] is continuous and

u, v : [c, d] → [a, b] are differentiable then
Z v (x)
d dv du
f (t) dt = f (v (x)) − f (u(x)) .
dx u(x) dx dx

Shripad M. Garge, IITB MA105: D1-08