Department of Mathematics

Indian Institute of Technology Guwahati

MA 101: Mathematics I
Differentiation
July-December 2019

The derivative of a function is usually tied up with the notion of the slope of the tangent
to the curve which represents the function, or with the rate of change of the function.
These interpretations are, in fact, the reason differentiation was introduced.

Definition 1. Let D ⊆ R and let x0 ∈ D be an interior point, that is, there exists some
δ > 0 such that (x0 − δ, x0 + δ) ⊆ D. A function f : D → R is said to be differentiable at
x0 if
f (x) − f (x0 ) f (x0 + h) − f (x0 )
lim or, equivalently lim
x→x0 x − x0 h→0 h
exists in R. If f is differentiable at x0 , then the derivative of f at x0 is

f (x) − f (x0 ) f (x0 + h) − f (x0 )

f 0 (x0 ) = lim = lim .
x→x0 x − x0 h→0 h
Further, f : D → R is said to be differentiable if f is differentiable at each x0 ∈ D.

In the following theorem, we prove that differentiability implies continuity.

Theorem 1. If f : D → R is differentiable at x0 ∈ D, then f is continuous at x0 .

Proof. For x 6= x0 , we write

f (x) − f (x0 )
f (x) = (x − x0 ) × + f (x0 ).
x − x0
f (x)−f (x0 )
Since lim (x − x0 ) = 0 and lim x−x0
= f 0 (x0 ), we find that
x→x0 x→x0

lim f (x) = f (x0 ).

x→x0

Therefore, f is continuous at x0 .

Example 1. f : R → R defined by f (x) = |x| is not differentiable at 0.

Solution. Let x 6= 0. Then

f (x) − f (0) |x| f (x) − f (0) |x|

lim = lim = 1 and lim = lim = −1.
x→0+ x−0 x→0+ x x→0− x−0 x→0− x

f (x)−f (0)
Therefore, lim x−0
does not exist, and hence f is not differentiable at 0.
x→0

x sin x1 if x 6= 0,

Example 2. Let f : R → R be defined by f (x) =
0 if x = 0.
Then f is not differentiable at 0.
 f (x)−f (0)
Solution. Since lim x−0
= lim sin x1 does not exist, f is not differentiable at 0.
x→0 x→0

x2 if x ∈ Q,
Example 3. f (x) = Then f : R → R is differentiable only at 0 and
0 if x ∈ R \ Q.
f 0 (0) = 0.

Solution. If x0 (6= 0) ∈ Q, then there exists a sequence (tn ) in R \ Q such that tn → x0 .

Since f (tn ) = 0 for all n ∈ N, f (tn ) → 0 6= x20 = f (x0 ). Hence f is not continuous at
x0 . Also, if u0 ∈ R \ Q, then there exists a sequence (rn ) in Q such that rn → u0 . Since
f (rn ) = rn2 → u20 6= 0 = f (u0 ), f is not continuous at u0 . Thus f is not continuous at any
x(6= 0) ∈ R and therefore f cannot be differentiable at any x(6= 0) ∈ R.
Again, for each ε > 0, choosing δ = ε > 0, we find that | f (x)−f x−0
(0)
| ≤ |x| < ε for all x ∈ R
f (x)−f (0)
satisfying 0 < |x| < δ. Hence lim x−0 = 0 and consequently f is differentiable at 0
x→0
with f 0 (0) = 0.

Theorem 2 (Rules for finding derivatives). Let D ⊆ R and let x0 ∈ D be an interior

point. Suppose f, g : D → R are differentiable at x0 . Then

(a) If α ∈ R, then the function αf is differentiable at x0 and (αf )0 (x0 ) = αf 0 (x0 ).

(b) The function f + g is differentiable at x0 and (f + g)0 (x0 ) = f 0 (x0 ) + g 0 (x0 ).

(c) (Product rule) The function f g is differentiable at x0 and (f g)0 (x0 ) = f 0 (x0 )g(x0 ) +
f (x0 )g 0 (x0 ).

(d) (Quotient rule) If g(x0 ) 6= 0, then the function f /g is differentiable at x0 and

0 f 0 (x0 )g(x0 ) − f (x0 )g 0 (x0 )

(f /g) (x0 ) = .
(g(x0 ))2

Theorem 3 (Chain rule for derivative). Let f : D → R and g : E → R. Let f (D) ⊆ E.

Suppose that f is differentiable at x0 and g is differentiable at f (x0 ). Then g ◦ f is
differentiable at x0 and (g ◦ f )0 (x0 ) = g 0 (f (x0 ))f 0 (x0 ).

Proof. We define a function h : E → R by

(
g(y)−g(f (x0 ))
y−f (x0 )
, y 6= f (x0 );
h(y) =
g 0 (f (x0 )), y = f (x0 ).

We have lim h(y) = g 0 (f (x0 )) = h(f (x0 )). Also, g(y) − g(f (x0 )) = h(y) × (y − f (x0 ))
y→f (x0 )
for all y ∈ E including y = f (x0 ). So, for x 6= x0 we have

(g ◦ f )(x) − (g ◦ f )(x0 ) g(f (x)) − g(f (x0 ))

=
x − x0 x − x0
f (x) − f (x0 )
= h(f (x))
x − x0
f (x)−f (x0 )
Hence, (g◦f )0 (x0 ) = lim h(f (x))× lim x−x0
= h(f (x0 ))f 0 (x0 ) = g 0 (f (x0 ))f 0 (x0 ).
x→x0 x→x0
 x2 sin x1 if x 6= 0,

Example 4. Let f : R → R defined by f (x) =
0 if x = 0.
Then f is differentiable at 0. But f 0 : R → R is not continuous at 0.
Solution. Clearly f is differentiable at all x (6= 0) ∈ R and f 0 (x) = 2x sin x1 − cos x1 for all
x (6= 0) ∈ R. Also, for each ε > 0, choosing δ = ε > 0, we find that | f (x)−f
x−0
(0)
| = |x sin x1 | ≤
|x| < ε for all x ∈ R satisfying 0 < |x| < δ. Hence lim f (x)−f
x−0
(0)
= 0 and consequently f is
x→0
differentiable at 0 with f 0 (0) = 0. Thus f : R → R is differentiable.
1
Again, since 2nπ → 0 but f 0 ( 2nπ
1
) → −1 6= f 0 (0), f 0 : R → R is not continuous at 0.
Definition 2. Let f : D → R. f has a local maximum at x0 ∈ D if there exists δ > 0
such that f (x) ≤ f (x0 ) for all x ∈ (x0 − δ, x0 + δ) ∩ D. Similarly, f has a local minimum
at x0 ∈ D if there exists δ > 0 such that f (x) ≥ f (x0 ) for all x ∈ (x0 − δ, x0 + δ) ∩ D.
Theorem 4. If f : D → R has a local maximum or local minimum at an interior point
x0 of D and if f is differentiable at x0 , then f 0 (x0 ) = 0.
Proof. Suppose that f has a local maximum at x0 , where x0 is an interior point of D.
Then there exists δ > 0 such that f (x) ≤ f (x0 ) for all x ∈ (x0 − δ, x0 + δ) ⊆ D. Hence,
f (x) − f (x0 ) f (x) − f (x0 )
≤ 0 for all x ∈ (x0 , x0 + δ) ⇒ lim ≤ 0;
x − x0 x→x 0 + x − x0
f (x) − f (x0 ) f (x) − f (x0 )
≥ 0 for all x ∈ (x0 − δ, x0 ) ⇒ lim ≥ 0.
x − x0 x→x0 − x − x0
Since f is differentiable at x0 , so f 0 (x0 ) = 0. If f has a local minimum at x0 then following
as above we readily obtain that f 0 (x0 ) = 0.
Theorem 5 (Rolle’s Theorem). Let f : [a, b] → R. Suppose that
(a) f is continuous on [a, b].
(b) f is differentiable on (a, b).
(c) f (a) = f (b).
Then there exists c ∈ (a, b) such that f 0 (c) = 0.
Proof. Since f is continuous on [a, b], so f has maximum and minimum values on [a, b]. If
both these extreme values are equal, then f is the constant function, and we can choose c
to be any point in (a, b). If f is not constant, then one of the extreme values is not equal
to f (a). Since f (a) = f (b), so f has a local maximum/minimum at a point c other than
a and b. In other words, c ∈ (a, b). Since f is differentiable at c, so f 0 (c) = 0.
Example 5. The equation x2 = x sin x + cos x has exactly two (distinct) real roots.
Solution. Let f (x) = x2 − x sin x − cos x for all x ∈ R. Then f : R → R is differentiable
(and hence continuous) with f 0 (x) = x(2 − cos x) for all x ∈ R. Since cos x 6= 2 for any
x ∈ R, the equation f 0 (x) = 0 has exactly one real root, namely x = 0. As a consequence
of Rolle’s theorem, it follows that the equation f (x) = 0 has at most two real roots. Also,
since f (−π) = π 2 + 1 > 0, f (0) = −1 < 0 and f (π) = π 2 + 1 > 0, by the intermediate
value property of continuous functions, the equation f (x) = 0 has at least one root in
(−π, 0) and at least one root in (0, π). Thus the equation f (x) = 0 has exactly two
(distinct) real roots and so the given equation has exactly two (distinct) real roots.
Example 6. Find the number of (distinct) real roots of the equation x4 +2x2 −6x+2 = 0.
Solution. Taking f (x) = x4 + 2x2 − 6x + 2 for all x ∈ R, we find that f : R → R is
twice differentiable with f 0 (x) = 4x3 + 4x − 6 and f 00 (x) = 12x2 + 4 for all x ∈ R. Since
f 00 (x) 6= 0 for all x ∈ R, as a consequence of Rolle’s theorem, it follows that the equation
f 0 (x) = 0 has at most one real root and hence the equation f (x) = 0 has at most two real
roots. Again, since f (0) = 2 > 0, f (1) = −2 < 0 and f (2) = 14 > 0, by the intermediate
value property of continuous functions, the equation f (x) = 0 has at least one real root
in (0, 1) and at least one real root in (1, 2). Therefore the given equation has exactly two
(distinct) real roots.
Theorem 6 (Mean Value Theorem (MVT)). If f : [a, b] → R is continuous and if f is
differentiable on (a, b), then there exists c ∈ (a, b) such that f (b) − f (a) = f 0 (c)(b − a).
Proof. Define the function g : [a, b] → R by
 
f (b) − f (a)
g(x) = f (x) − f (a) + (x − a) .
b−a

Clearly g satisfies the conditions of Rolle’s theorem. Therefore, there exists c ∈ (a, b) such
that g 0 (c) = 0. This gives f (b) − f (a) = f 0 (c)(b − a).
Theorem 7. Let I be an interval and let f : I → R be differentiable. Then
(a) f 0 (x) = 0 for all x ∈ I ⇔ f is constant on I.

(b) f 0 (x) ≥ 0 for all x ∈ I ⇔ f is increasing on I.

(c) f 0 (x) ≤ 0 for all x ∈ I ⇔ f is decreasing on I.

(d) f 0 (x) > 0 for all x ∈ I ⇒ f is strictly increasing on I.

(e) f 0 (x) < 0 for all x ∈ I ⇒ f is strictly decreasing on I.

(f) f 0 (x) 6= 0 for all x ∈ I ⇒ f is one-one on I.

Proof of (a). If f is constant on I then it easily follows that f 0 (x) = 0 for all x ∈ I.
Conversely, suppose that f 0 (x) = 0 for all x ∈ I. Let x1 , x2 ∈ I be such that x1 < x2 . By
MVT, there exists c ∈ (x1 , x2 ) such that f (x2 )−f (x1 ) = f 0 (c)(x2 −x1 ). Since c ∈ (x1 , x2 ),
f 0 (c) = 0 and therefore f (x1 ) = f (x2 ). This proves that f is a constant function.
Proof of (b) and (c). Suppose that f 0 (x) ≥ 0 for all x ∈ I. Let x1 , x2 ∈ I be such that
x1 < x2 . Then applying MVT to [x1 , x2 ], there exists some c ∈ (x1 , x2 ) such that

f (x2 ) − f (x1 ) = f 0 (c)(x2 − x1 ) ≥ 0.

This gives f (x2 ) ≥ f (x1 ), and hence f is increasing. Conversely, suppose that f is
differentiable and increasing on I. Let c ∈ I. Then for any x 6= c, we have
f (x) − f (c) f (x) − f (c)
≥ 0, and hence f 0 (c) = lim ≥ 0.
x−c x→c x−c
The proof of (c) follows similarly.
Proof of (d) and (e). Suppose that f 0 (x) > 0 for all x ∈ I. Let x1 , x2 ∈ I be such that
x1 < x2 . Then applying MVT to [x1 , x2 ], there exists some c ∈ (x1 , x2 ) such that

f (x2 ) − f (x1 ) = f 0 (c)(x2 − x1 ) > 0.

This gives f (x2 ) > f (x1 ), and hence f is strictly increasing. The proof of (e) follows
similarly.
Proof of (f ). If f is not injective, there would be two points x1 , x2 ∈ I such that x1 < x2
and f (x1 ) = f (x2 ). Applying MVT to [x1 , x2 ], we would then conclude that there is a
f (x2 ) − f (x1 )
c ∈ (x1 , x2 ) ⊆ I such that f 0 (c) = = 0. This contradicts the hypothesis
x2 − x1
that f 0 (x) 6= 0 for all x ∈ I.

Remark 1. The converse of (d) and (e) are not true in general. For example, f : R → R
defined by f (x) = x3 is strictly increasing but f 0 (0) = 0.

Example 7. We have sin x ≤ x for all x ≥ 0.

Proof. Let f (x) = sin x − x for all x ∈ R. Then f 0 (x) = cos x − 1 for all x ∈ R. Since
f 0 (x) ≤ 0 for all x ∈ R, so f is decreasing on R. Hence f (x) ≤ f (0) = 0 for all x ≥ 0.
Therefore sin x ≤ x for all x ≥ 0.
x3
Example 8. sin x ≥ x − 6
for all x ∈ [0, π2 ].
3
Solution. Let f (x) = sin x − x + x6 for all x ∈ [0, π2 ]. Then f : [0, π2 ] → R is infinitely
2
differentiable and f 0 (x) = cos x − 1 + x2 , f 00 (x) = sin x + x and f 000 (x) = 1 − cos x for
all x ∈ [0, π2 ]. Since f 000 (x) ≥ 0 for all x ∈ [0, π2 ], f 00 is increasing on [0, π2 ]. Hence
f 00 (x) ≥ f 00 (0) = 0 for all x ∈ [0, π2 ]. This shows that f 0 is increasing on [0, π2 ] and so
f 0 (x) ≥ f 0 (0) = 0 for all x ∈ [0, π2 ]. Thus f is increasing on [0, π2 ] and so f (x) ≥ f (0) = 0
3
for all x ∈ [0, π2 ]. Therefore sin x ≥ x − x6 for all x ∈ [0, π2 ].

Example 9. If f (x) = x3 + x2 − 5x + 3 for all x ∈ R, then f is one-one on [1, 5] but not

one-one on R.

Proof. f : R → R is differentiable with f 0 (x) = 3x2 + 2x − 5 for all x ∈ R. Clearly

f 0 (x) 6= 0 for all x ∈ (1, 5) and hence f is one-one on [1, 5]. Again, since f (0) = 3,
f (1) = 0 and f (2) = 5, by the intermediate value property of continuous functions, there
exist x1 ∈ (0, 1) and x2 ∈ (1, 2) such that f (x1 ) = 1 = f (x2 ). Therefore f is not one-one
on R.

Theorem 8 (L’Hôpital’s rules). Let f : (a, b) → R and g : (a, b) → R.

(1) Suppose that f and g are differentiable at x0 ∈ (a, b), f (x0 ) = g(x0 ) = 0 and
0 (x )
g 0 (x0 ) 6= 0. Then lim fg(x)
(x)
= fg0 (x 0
0)
.
x→x0

(2) Suppose that f : (a, b) → R and g : (a, b) → R are differentiable such that
0 (x)
lim f (x) = lim g(x) = 0 and g 0 (x) 6= 0 for all x ∈ (a, b). If lim fg0 (x) = `,
x→a+ x→a+ x→a+
then lim fg(x)
(x)
= `.
x→a+
 √
1+x−1
Example 10. lim x
= 12 .
x→0

Solution. Applying (first version of) L’Hôpital’s rule, we obtain

√ d
√
1+x−1 dx
( 1 + x − 1)|x=0 1
lim = d
= .
x→0 x dx
(x)|x=0 2

1−sin x
Example 11. limπ = 41 .
x→ 2 1+cos 2x

Solution. Applying L’Hôpital’s rule twice, we obtain

1 − sin x − cos x sin x 1
limπ = limπ = limπ = .
x→ 2 1 + cos 2x x→ 2 −2 sin 2x x→ 2 −4 cos 2x 4

x2 sin x1
Example 12. lim = 0.
x→0 sin x

Solution. For all x(6= 0) ∈ R, we have 0 ≤ |x sin x1 | ≤ |x|. Since lim |x| = 0, by sandwich
x→0
theorem (for limit of functions), we get lim |x sin x1 | = 0 and hence lim x sin x1 = 0. It
x→0 x→0
x2 sin 1 x sin 1 lim x sin x1
x→0 0
follows that lim sin x x = lim sin x
x
= sin x
lim
= 1
= 0.
x→0 x→0 x x→0 x

x−sin x
Example 13. lim = 21 .
x→∞ 2x+sin x

Solution. Since | sinx x | ≤ 1

x
for all x > 0 and since lim 1
= 0, we get lim sin x
= 0.
x→∞ x x→∞ x
Consequently
sin x
x − sin x 1− x 1
lim = lim sin x
= .
x→∞ 2x + sin x x→∞ 2 + 2
x

A very useful technique in the analysis of real functions is the approximation of func-
tions by polynomials.
Theorem 9 (Taylor’s theorem). Let f : [a, b] → R be such that f and its derivatives
f 0 , f 00 , . . . , f (n) are continuous on [a, b] and that f (n+1) exists on (a, b). If x0 ∈ [a, b], then
for any x ∈ [a, b] there exists a point c between x and x0 such that

f 00 (x0 ) f (n) (x0 )

f (x) = f (x0 ) + f 0 (x0 )(x − x0 ) + (x − x0 )2 + · · · + (x − x0 )n
2! n!
f (n+1) (c)
+ (x − x0 )n+1
(n + 1)!
= Pn (x) + Rn (x),
f 00 (x0 ) f (n) (x0 )
where Pn (x) = f (x0 ) + f 0 (x0 )(x − x0 ) + 2!
(x − x0 )2 + · · · + n!
(x − x0 )n and
f (n+1) (c)
Rn (x) = (n+1)! (x − x0 )n+1 . The polynomial Pn (x) is called the nth Taylor polynomial of
f and Rn (x) is called the remainder.
Proof. Let x and x0 be given and let J denote the closed interval with endpoints x0 and
x. We define the function F on J by
f (n) (t)
F (t) := f (x) − f (t) − f 0 (t) · (x − t) − · · · − (x − t)n , (1)
n!
where t ∈ J. We have
(x − t)n (n+1)
F 0 (t) = − f (t).
n!
If we define G on J by
 n+1
x−t
G(t) := F (t) − F (x0 )
x − x0
for t ∈ J, then G(x0 ) = G(x) = 0. Applying Rolle’s theorem, there exists c between x
and x0 such that
(x − c)n
0 = G0 (c) = F 0 (c) + (n + 1) F (x0 ).
(x − x0 )n+1
Hence, we obtain
1 (x − x0 )n+1 0
F (x0 ) = − F (c)
n + 1 (x − c)n
1 (x − x0 )n+1 (x − c)n n+1 f (n+1) (c)
= n
f (c) = (x − x0 )n+1 . (2)
n + 1 (x − c) n! (n + 1)!
From (1) and (2) we readily obtain the required result.
2 √
Example 14. We have 1 + x2 − x8 ≤ 1 + x ≤ 1 + x2 for all x > 0.
√
Solution. Let x > 0 and let f (t) = 1 + t for all x ∈ [0, x]. Then f : [0, x] → R is twice
differentiable and f 0 (t) = 2√11+t , f 00 (t) = − 4(1+t)
1
3/2 for all t ∈ [0, x]. By Taylor’s theorem,
2 x2
there exists c ∈ (0, x) such that f (x) = f (0) + xf 0 (0) + x2! f 00 (c) = 1 + x
2
− 8
· 1
(1+c)3/2
.
1 x x2
√
Since 0 < (1+c) 3/2 < 1, we get 1 + 2 − 8 ≤ 1 + x ≤ 1 + x2 .
Theorem 10 (Results on local maxima and local minima). Let x0 ∈ (a, b) and let n ≥ 2.
Also, let f, f 0 , ..., f (n) be continuous on (a, b) and f 0 (x0 ) = f 00 (x0 ) = · · · = f (n−1) (x0 ) = 0
but f (n) (x0 ) 6= 0.
(a) If n is even and f (n) (x0 ) < 0, then f has a local maximum at x0 .
(b) If n is even and f (n) (x0 ) > 0, then f has a local minimum at x0 .
(c) If n is odd, then f has neither a local maximum nor a local minimum at x0 .
Proof. Applying Taylor’s theorem at x0 , we find that for x ∈ (a, b) we have
f (n) (c)
f (x) = Pn−1 (x) + Rn−1 (x) = f (x0 ) + (x − x0 )n ,
n!
where c is some point between x0 and x. Since f (n) is continuous and f (n) (x0 ) 6= 0, so there
exists δ > 0 such that f (n) (x) will have the same sign as f (n) (x0 ) for all x ∈ (x0 −δ, x0 +δ).
If x ∈ (x0 − δ, x0 + δ), then the point c ∈ (x − x0 , x + x0 ) and consequently f (n) (c) and
f (n) (x0 ) will have the same sign.
 (a) If n is even and f (n) (x0 ) < 0, then we have f (n) (c) < 0 and (x − x0 )n ≥ 0 so that
Rn−1 (x) ≤ 0 for all x ∈ (x0 −δ, x0 +δ). Hence f (x) ≤ f (x0 ) for all x ∈ (x0 −δ, x0 +δ),
and therefore f has a local maximum at x0 .

(b) If n is even and f (n) (x0 ) > 0, then we have f (n) (c) > 0 and (x − x0 )n ≥ 0 so that
Rn−1 (x) ≥ 0 for all x ∈ (x0 −δ, x0 +δ). Hence f (x) ≥ f (x0 ) for all x ∈ (x0 −δ, x0 +δ),
and therefore f has a local minimum at x0 .

(c) If n is odd then (x − x0 )n is positive if x > x0 and negative if x < x0 . Consequently,

if x ∈ (x0 − δ, x0 + δ) then Rn−1 (x) will have opposite signs to the left and to the
right of x0 . Therefore f has neither a local maximum nor a local minimum at x0 .

Theorem 11. If f (x) = x5 − 5x4 + 5x3 + 12 for all x ∈ R, then f has a local maximum
only at 1 and a local minimum only at 3.
Proof. f : R → R is infinitely differentiable and f 0 (x) = 5x2 (x − 1)(x − 3), f 00 (x) =
10x(2x2 − 6x + 3), f 000 (x) = 30(2x2 − 4x + 1) for all x ∈ R. Since f 0 (x) = 0 iff x = 0, 1,
or 3, f has neither a local maximum nor a local minimum at any point of R \ {0, 1, 3}.
Again, since f 00 (1) = −10 < 0, f 00 (3) = 90 > 0, f 00 (0) = 0 and f 000 (0) = 30 6= 0, f has a
local maximum at 1 (with local maximum value f (1) = 13), f has a local minimum at 3
(with local minimum value f (3) = −15) and f has neither a local maximum nor a local
minimum at 0.
∞
an (x − x0 )n , where x0 , an ∈ R
P
Definition 3. A power series is a series of the form
n=0
for n = 0, 1, 2, . . . and x ∈ R.
Remark 2. Since the transformation X = x − c reduces a power series around c to a
∞
an x n .
P
power series around 0, it is sufficient to consider the series
n=0
∞ ∞ ∞
xn
n!xn xn
P P P
Example 15. (a) n!
(b) (c)
n=0 n=0 n=0

Theorem 12. We have

∞
an xn converges for x = x1 6= 0, then it converges absolutely for all x ∈ R
P
(a) If
n=0
satisfying |x| < |x1 |.
∞
an xn diverges for x = x2 , then it diverges for all x ∈ R satisfying |x| > |x2 |.
P
(b) If
n=0

Proof. Observe that (a) ⇔ (b). So we prove part (a) only.

∞
X
(a) Let x1 6= 0 and put yn = an xn1 for all n ≥ 0. It is given that yn is convergent.
n=0
Hence, there exists M > 0 such that |yn | = |an xn1 | < M . Let x ∈ R be such that
|x| < |x1 |. Let r = |x|x|1 | . Now,

|an xn | = |an xn1 |rn < M rn for all n.

 ∞
X ∞
rn converges, and by comparison test an xn converges
P
Since 0 ≤ r < 1, so
n=0 n=0
absolutely.

∞
an xn , there exists a
P
Definition 4 (Radius of convergence). For every power series
n=0
unique R satisfying 0 ≤ R ≤ ∞ such that the series converges absolutely if |x| < R and
diverges if |x| > R.
∞ p 1
an xn . Let β = lim sup
P
Theorem 13. Consider the power series n
|an | and R = β
n=0
(we define R = 0 if β = ∞ and R = ∞ if β = 0). Then
∞
an xn converges absolutely for |x| < R
P
(a)
n=0

∞
an xn diverges for |x| > R.
P
(b)
n=0

(c) No conclusion if |x| = R.

p p ∞ ∞
Proof. Let xn = an xn . Then an x n
P P
n
|xn | = |x| n |an |. By Root test, xn =
p p n=0 n=0
converges absolutely if lim sup n |xn | = |x| lim sup n |an | = |x|β < 1. Equivalently,
∞ ∞
an xn converges absolutely for |x| < β1 = R. Again, by Root test an xn diverges for
P P
n=0 n=0
|x| > R.
∞  
an xn . Suppose β = lim  an+1 1
P
Theorem 14. Consider the power series and R =
 
an  β
n=0
(We define R = 0 if β = ∞ and R = ∞ if β = 0). Then
∞
an xn converges for |x| < R
P
(a)
n=0

∞
an xn diverges for |x| > R.
P
(b)
n=0

(c) No conclusion if |x| = R.

Proof. The proof readily follows using the Ratio test.

∞
xn
P
Example 16. The power series n2
converges if and only if x ∈ [−1, 1].
n=0

Solution. If x = 0, then the given series becomes 0 + 0 + · · · , which is clearly convergent.

n
Let x (6= 0) ∈ R and let an = xn2 for all n ∈ N. Then lim | an+1 | = |x|. Hence by ratio
n→∞ an
∞
P
test, an is convergent (absolutely) if |x| < 1, i.e. if x ∈ (−1, 1) and is not convergent
n=1
if |x| > 1, i.e. if x ∈ (−∞, −1) ∪ (1, ∞). Therefore the radius of convergence of the
 ∞ ∞
1
P P
given power series is 1. Again, if |x| = 1, then |an | = n2
is convergent and hence
n=1 n=1
∞
P
an is also convergent. Therefore the interval of convergence of the given power series
n=1
is [−1, 1].
∞
(−1)n
− 1)n , the radius of convergence is 4 and
P
Example 17. For the power series n·4n
(x
n=1
the interval of convergence is (−3, 5].
Solution. Method-I: If x = 1, then the given series nbecomes 0 + 0 + · · · , which is
clearly convergent. Let x(6= 1) ∈ R and let an = (−1) n.4n
(x − 1)n for all n ∈ N. Then
∞
lim | an+1 1
an is convergent (absolutely) if 14 |x−1| < 1,
P
an
| = 4
|x−1|. Hence by ratio test,
n→∞ n=1
i.e. if x ∈ (−3, 5) and is not convergent if 41 |x − 1| > 1, i.e. if x ∈ (−∞, −3) ∪ (5, ∞).
Therefore the radius of convergence of the given power series is 4. Again, if x = −3, then
∞ ∞ ∞ ∞
P P 1
P P (−1)n
an = n
is not convergent. If x = 5, then a n = n
is convergent by Leibniz
n=1 n=1 n=1 n=1
test. Therefore the interval of convergence of the given power series is (−3, 5].
n
Method-II: Let an = (−1) n·4n
for all n ∈ N. Clearly, β = lim | an+1 an
| = 14 . Therefore
the radius of convergence of the given power series is 4. Thus the given series is conver-
gent (absolutely) if |x − 1| < 4, that is, if x ∈ (−3, 5) and is not convergent if |x − 1| > 4,
that is, if x ∈ (−∞, −3) ∪ (5, ∞). Rest of the proof is similar.
Theorem 15 (Term by term differentiation of power series). A power series can be
∞
X
differentiated term by term within the interval of convergence. In fact, if an xn has
n=0
∞
X
radius of convergence R, then f 0 (x) = nan xn−1 for |x| < R.
n=1
p 1 p p
Proof. It is clear that lim sup n |nan | = lim sup{n n n |an |} = lim sup n |an | = β. Thus,
X∞ X ∞
n
both the series an x and nan xn−1 have the same radius of convergence. However,
n=0 P n=1 n−1
to prove that the series ∞ n=1 nan x converges to f 0 (x) requires another concept called
uniform convergence which is beyond the scope of this course.
Definition 5. If a function f has derivatives of all orders at a point c ∈ R, then we can
(n)
calculate the Taylor coefficients by a0 = f (c), an = f n!(c) for all n ∈ N. In this way, we
∞ ∞
X f (n) (c) X f (n) (c)
obtain a power series (x − c)n . The power series (x − c)n converges
n=0
n! n=0
n!
to f (x) for |x − c| < R if and only if the sequence (Rn (x)) of remainders converges to 0
∞
X f (n) (c)
for each x in |x − c| < R. The power series (x − c)n is called the Taylor series
n=0
n!
of f at c. The Taylor series of a function f at c = 0 is known as Maclaurin’s series.
∞ 2n−1
n+1 x
X
Example 18. The Maclaurin series (−1) of sin x converges to sin x for
n=1
(2n − 1)!
all x ∈ R.
Solution. If f (x) = sin x for all x ∈ R, then f : R → R is infinitely differentiable and
f (2n−1) (x) = (−1)n+1 cos x, f (2n) (x) = (−1)n sin x for all x ∈ R and for all n ∈ N. Hence
∞
x2n−1
(−1)n+1 (2n−1)!
P
the Maclaurin series for sin x is the series , where x ∈ R. For x = 0, the
n=1
Maclaurin series of sin x becomes 0 − 0 + 0 − · · · , which clearly converges to sin 0 = 0. Let
x(6= 0) ∈ R. The remainder term in the Taylor expansion of sin x about the point 0 is
xn+1 (n+1)
given by Rn (x) = (n+1)! f (cn ), where cn lies between 0 and x. Since | sin cn | ≤ 1 and
|x|n+1 |x|n+2 (n+1)! |x|
| cos cn | ≤ 1, we get |Rn (x)| ≤ (n+1)!
. Also, since lim(n+2)!
· |x|n+1
= lim = 0 < 1,
n→∞ n→∞ n+2
n+1
|x|
we get lim (n+1)! = 0 and hence it follows that lim Rn (x) = 0. Therefore the Maclaurin
n→∞ n→∞
series of sin x converges to sin x.

Example 19. Let f : R → R be defined by

 −1/x2
e if x 6= 0,
f (x) =
0 if x = 0.

Then f is differentiable any number of times, and that f (n) (0) = 0 for every n ∈ N.
Hence, in a neighborhood of 0, the Maclaurin series of f is identically zero. But the
function takes nonzero value at x 6= 0. Thus, the Maclaurin series of f does not converge
to f . The reason is that the remainder term Rn (c) does not converge to 0 for any c 6= 0.
Therefore, an infinitely differentiable function may not have Taylor series representation.