Académique Documents
Professionnel Documents
Culture Documents
Expansion of Functions
7 )
Theorem on Local Extrema If f (c) is a local extremum, then either f (x) is not
1
20
differentiable at c or f ′ (c) = 0. That is, at a local max or min f either has no tangent,
or f (x) has a horizontal tangent there.
(
if : [a, b] → R is differentiable on (a, b)
Rolle’s Theorem Let a < b and suppose
c h
and continuous on [a, b], and f (a) = fg(b). Then there exists a c ∈ (a, b) such that
′
B a f has a horizontal tangent somewhere
f (c) = 0. That is, under these hypotheses,
between a and b. S
Proof We seek a c in (a, b) with f ′ (c) = 0. That is, we wish to show that f has a
horizontal tangent somewhere between a and b.
Since f is continuous on the closed interval [a, b], the Extreme Value Theorem says
that f has a maximum value f (M ) and a minimum value f (m) on the closed interval
[a, b]. Either f (M ) = f (m) or f (M ) 6= f (m).
Case 1. We suppose the maximum value f (M ) = f (m), the minimum value. So all
values of f on [a, b] are equal, and f is constant on [a, b]. Then f ′ (x) = 0 for all x in
(a, b). So one may take c to be anything in (a, b); for example, c = a+b 2
would suffice.
Case 2. Now we suppose f (M ) 6= f (m). So at least one of f (M ) and f (m) is not
equal to the value f (a) = f (b).
Case 2.a We first consider the case where the maximum value f (M ) 6= f (a) = f (b).
So M is neither a nor b. But M is in [a, b] and not at the end points. So M must
be in the open interval (a, b). We have the maximum value f (M ) ≥ f (x) for all x
in the closed interval [a, b] which contains the open interval (a, b). So we also have
f (M ) ≥ f (x) for every x in the open interval (a, b), Since M is also in the open
interval (a, b), this means by definition that f (M ) is a local maximum.
Since M is in the open interval (a, b), by hypothesis we have that f is differentiable at
M . Now by the Theorem on Local Extrema, we have that f has a horizontal tangent
1
at m; that is, we have that f (M ) = 0. So we take c = M , and we are done with this
case.
Case 2.b We now consider the case where the minimum value f (m) 6= f (a) = f (b).
(This case is very similar to the previous case.)
So m is neither a nor b. But m is in [a, b] and not at the endpoints. So m must be
in the open interval (a, b). We have the minimum value f (m) ≤ f (x) for all x in
the closed interval [a, b] which contains the open interval (a, b). Thus f (m) ≤ f (x)
for every x in the open interval (a, b), Since m is also in the open interval (a, b), this
means by definition that f (m) is a local minimum.
Since m is in the open interval (a, b), by hypothesis we have that f is differentiable at
m. Now by the Theorem on Local Extrema, we have that f has a horizontal tangent
at m; that is, we have that f ′ (m) = 0. So we take c = m, and we are done with this
case.
This completes the proof.
Langrange’s Mean Value Theorem “MVT” Suppose f : [a, b] → R is differen-
7 )
such that f ′ (c) = f (b)−f (a)
0 1
tiable on (a, b), and continuous on [a, b] where a < b. Then there exists a c ∈ (a, b)
b−a
.
( 2
Proof The equation of the secant through i(a, f (a)) and (b, f (b)) is
g ch (x − a)
a
y − f (a) =
B
f (b)−f (a)
(b−a)
Let
h i
f (b)−f (a)
g(x) = f (x) − (b−a)
(x − a) + f (a)
Note that g(a) = g(b) = 0. Also, g is continuous on [a, b] and differentiable on (a, b)
since f is. So by Rolle’s Theorem there exists c in (a, b) such that g ′ (c) = 0.
f (b)−f (a) f (b)−f (a)
But g ′ (x) = f ′ (x) − b−a
, so g ′ (c) = f ′ (c) − b−a
= 0.
f (b)−f (a)
Therefore, f ′ (c) = b−a
and the proof is complete.
Cauchy’s Mean Value Theorem Let f, g : [a, b] → R be continuous and differen-
tiable on (a, b). Then there exists c ∈ (a, b) such that
f ′ (c)(g(b) − g(a)) = g ′ (c)(f (b) − f (a)).
Proof Hint: Define h(x) = f (x) − rg(x), where r is fixed in such a way that
h(a) = h(b), namely
f (b) − f (a)
r= .
g(b) − g(a)
2
Then use Rolle’s Theorem on h(x).
Generalization for determinants: Assume that f (x), g(x), and h(x) are differen-
tiable functions on (a, b) that are continuous on [a, b]. Define
and if we place h(x) = 1, we get Cauchy’s mean value theorem. If we place h(x) = 1
and g(x) = x we get Lagrange’s mean value theorem.
The proof of the generalization is quite simple: Each of D(a) and D(b) are determi-
7 )
nants with two identical rows, hence D(a) = D(b) = 0. The Rolle’s theorem implies
1
20
that there exists c ∈ (a, b) such that D′ (c) = 0.
(
Taylor Polynomial for f (x) about a: Suppose f (x) and its derivatives f ′ (x),
i
by the formula. g ch
f ′′ (x), . . . , f (n) (x) exist at a. We define the nth Taylor polynomial for f (x) about a
a
SB
Tn (x) = f (a) + f ′ (a)(x − a) + f ′′ (a)
(x − a)2
2!
+ · · · + f (n) (a)
(x − a)n
n!
i.e. n
X (x − a)k
Tn (x) = f (k) (a) ,
k=0
n!
where f (0) (x) = f (x) and f (k) (x) denote the k th derivative of f (x).
Taylor’s Theorem Let n ∈ N, let I := [a, b], and let f (x) : I → R be such that f
and its derivatives f ′ (x), f ′′ (x), . . . , f (n) (x) are continuous on I and f (n+1) (x) exists
on (a, b). If x0 ∈ I, then for any x ∈ I there exists a point c between x and x0 such
that
′′
′ f (x0 ) f (n) (x0 ) f (n+1) (c)
f (x) = f (x0 )+f (x0 )(x−x0 )+ (x−x0 )2 +· · ·+ (x−x0 )n + (x−x0 )n+1 .
2! n! (n + 1)!
i.e. f (x) = Tn (x) + Rn (x). This is the Lagrange form of the remainder.
For the special case x0 = 0 the Taylor Theorem becomes
′′
′ f (0) 2 f (n) (0) n f (n+1) (c) n+1
f (x) = f (0) + f (0)x + x + ··· + x + x .
2! n! (n + 1)!
3
Example: Use Taylor’s Theorem to prove that 1 − 12 x2 ≤ cos x for all x ∈ R.
′′′
f (c) 3 sin c 3
R2 (x) = x = x.
3! 6
if 0 ≤ x ≤ π, then 0 ≤ c < π; since c and x3 are both positive, we have R2 (x) ≥ 0.
Also, if −π ≤ x ≤ 0, then −π ≤ c ≤ 0; since sin c and x3 are both negative, we
again have R2 (x) ≥ 0.Therefore, we see that 1 − 12 x2 ≤ cos x for | x |≤ π. If | x |≥ π,
then we have 1 − 21 x2 < −3 ≤ cos x and the inequality is trivially valid. Hence, the
inequality holds for all x ∈ R.
7 )
Example: For any k ∈ N, and for all x > 0, we have
1
1 1 1 ( 201
i
x − x2 + · · · − x2k < ln(1 + x) < x − x2 + · · · + x2k+1 .
2 2k 2
g ch 2k + 1
a
SB
1
Using the fact that the derivative of ln(1 + x) is (1+x)
for x > 0, we see that the nth
Taylor polynomial for ln(1 + x) with x0 = 0 is
1 1
Pn (x) = x − x2 + · · · + (−1)n−1 xn
2 n
and the remainder is given by
4
√
For example, if we let x = 0.3, we get the approximation P2 (0.3) = 1.09 for 3 1.3.
Moreover, since c > 0 in this case, then (1 + c)−8/3 < 1 and so the error is at most
5 3 3 1
R2 (0.3) ≤ ( ) = < 0.17 ∗ (10)−2 .
81 10 600
√
Hence, we have | 3
1.3 − 1.09 |< 0.5 ∗ 10−2 , so that two decimal place accuracy is
assured.
Example: Prove that the equation x7 + x5 + x3 + 1 = 0 has exactly one real solution.
You should use Rolle’s Theorem at some point in the proof.
Let y = y(x) = x7 + x5 + x3 + 1. y(0) = 1, y(−1) = −2. By EVT, there exists at least
one real root in x ∈ (−1, 0) such that y(x) = 0. Now I claim that there is EXACTLY
one such real root, by using the method of contradiction.
Suppose not, there exists at least 2 real roots x1 , x2 such that y(x1 ) = 0, y(x2 ) = 0.
Since y is differentiable, by Rolle’s theorem, there exists a number a ∈ (x1 , x2 ) such
that y ′ (a) = 0. However, y ′ (x) = 7x6 + 5x4 + 3x2 > 0 for all x 6= 0.
7 )
1
Problem Set
( 20
i
ch
1. Use Rolle’s theorem to prove that the equation 5x3 − 2x2 + x − 6 = 0 can not
g
have more than one real root.
a
SB
2. Prove that the equation x13 + 7x3 − 5 = 0 has exactly one real root.
6. In each of the following, give an example of a function that fits the given con-
ditions and for which the conclusion of Rolle’s theorem does not hold:
8. Find the 4th Taylor polynomial of the function f (x) = sin x about 0.
5
9. Find the 4th Taylor polynomial of f (x) = 3 + 5x2 − 4x3 + x4 about 0.
10. Find the 4th Taylor polynomial of f (x) = 3 + 5x2 − 4x3 + x4 about 1.
11. Find the 4th Taylor polynomial of the function f (x) = ex about 0.
√ 1 1
12. Use Mean Value Theorem to show that 3 28 lies between 3 + 28 and 3 + 27
.
x2 x3 x x2 x3 x
1+x+ + <e <1+x+ + e .
2! 3! 2! 3!
14. If f (x) and g(x) are differentiable functions for 0 ≤ x ≤ 1 such that f (0) = 5,
g(0) = 1, f (1) = 8 and g(1) = 2, then show that there exists c satisfying
0 < c < 1 and f ′ (c) = 3g ′ (c).
g c
f (a) = f (b) = 0. Prove that there his a point c ∈ (a, b) such that g (c)f (c) +
′
′
f (c) = 0. a Hint: Let h(x) = e .f (x) g(x)
S Bshow that
18. Using Mean Value Theorem,
x−1
(a) x
< log x < x − 1 for x > 1.
x
(b) e ≥ 1 + x for x ∈ R.
x2
(c) 1 − 2!
< cos x, for x 6= 0.
x3
(d) x − 3!
< sin x, for x > 0.
(e) x > sin x for 0 < x < π2 .
x
(f) 1+x
< log(1 + x) < x for all x > 0.
19. (Using the MVT) Prove that | sin x − sin y| ≤ |x − y| for all x, y ∈ R. Conse-
quently, ∀x ∈ R, | sin x| ≤ |x|.
20. Let f : [a, b] → R be continuous on [a, b] and differentiable on (a, b). Suppose
that f (a) = a and f (b) = b. Show that there is c ∈ (a, b) such that f ′ (c) = 1.
Further, show that there are distinct c1 , c2 ∈ (a, b) such that f ′ (c1 ) + f ′ (c2 ) = 2.
*******
6
Radius of Curvature
1. The tangent ∆ψ through which the tangent turns as the point of contact travels
from one end to the other of the arcP Q is called the total curvature of the
arcP Q.
∆ψ
2. The mean or average curvature of the arc P Q is defined as the ratio ∆s
.
3. The curvature (k) at a point P of the curve is defined as the limiting value of
mean curvature when the arc ∆s → 0; that is
∆ψ dψ
Curvature (k) at P = lim = .
7 ) ∆s→0 ∆s ds
0
4. Suppose that k 6= 0. Take a quantity ρ such that 1 ρ= = . 1 ds
h
Now construct a circle of radius ρ and centre
have the same tangent at P . The circlecis
idrawn in such a way that it lies on the
C
a
same side of the tangent as the curve. gThis circle has the same curvature as the given
S Bthe circle of curvature at P ; its centre C is the
curve at P . We call this circle as
centre of the curvature for the curve at P ; its radius CP (= ρ), normal to the curve,
is the radius of curvature of the curve at P .
Formula for the radius of curvature:
7
3. Polar equation r = f (θ):
3
(r2 + r12 ) 2 dr d2 r
ρ= 2 , where r 1 = and r 2 = .
r + 2r12 − rr2 dθ dθ2
Problem Set
1. Show that the radius of curvature at a point of the curve x = aeθ (sin θ −
cos θ), y = aeθ (sin θ + cos θ) is twice the distance of the tangent at that point
from the origin.
2. Show that for the cycloid x = a(θ − sin θ), y = a(1 − cos θ), the radius of
curvature at any point is twice the portion of the normal intercepted between
the curvature and the axis of x.
7 )
1
4. Show that the radius of curvature of a logarithmic spiral r = aeθ cot α is propor-
tional to the radius vector.
( 20
5. For the catenary y = c cosh find thei radius of curvature at any point (x, y).
h
x
Also show that the length of thecnormal at (x, y) to the curve intercepted
g
c
9. The tangents at two points P and Q on the cycloid x = a(θ − sin θ), y =
a(1 − cos θ) are at right angles. Show that if ρ1 and ρ2 be the radii of curvature
at the points, then ρ21 + ρ22 = 16a2 .
10. Show that the radius of curvature ρ at any point of the cycloid s2 = 8ay is given
y 21
by ρ = 4a(1 − 2a ) .
2 2 2
11. Prove that the radius of curvature at any point (x, y) of the curve x 3 + y 3 = a 3
is three times the length of the perpendicular from the origin to the tangent at
(x, y).
*******
8
Asymptotes
7 )
II. To find the asymptotes parallel to y-axis, equate to zero the coefficient of the
1
20
highest power of y in the equation, provided this is not merely a constant.
i (
III. Let f (x, y) = 0 be the equation of any rational algebraic curve of the nth degree.
ch
Arranging this equation in groups of homogeneous terms in x and y, we get
g
a
(a0 xn + a1 xn−1 y + a2 xn−2 y 2 + · · · + an y n ) + (b0 xn−1 + b1 xn−2 y + b2 xn−2 y 2 + · · · +
SB
bn−1 y n−1 ) + (c0 xn−2 + c1 xn−3 y + c2 xn−4 y 2 + · · · + cn−2 y n−2 ) + (d0 xn−3 + d1 xn−4 y +
d2 xn−5 y 2 + · · · + dn−3 y n−3 ) + · · · + K = 0.
Let φn (x, y) = a0 xn + a1 xn−1 y + a2 xn−2 y 2 + · · · + an y n ,
φn−1 (x, y) = b0 xn−1 + b1 xn−2 y + b2 xn−2 y 2 + · · · + bn y n ,
φn−2 (x, y) = c0 xn−2 + c1 xn−3 y + c2 xn−4 y 2 + · · · + cn−2 y n−2 ,
φn−3 (x, y) = d0 xn−3 + d1 xn−4 y + d2 xn−5 y 2 + · · · + dn−3 y n−3 , and so
on.
Putting x = 1 and y = m, we get
φn (m) = a0 + a1 m + a2 m2 + · · · + an mn ,
φn−1 (m) = b0 + b1 m + b2 m2 + · · · + bn−1 mn ,
φn−2 (m) = c0 + c1 m + c2 m2 + · · · + cn−2 mn−2 ,
φn−3 (m) = d0 + d1 m + d2 m2 + · · · + dn−3 mn−3 , and so on.
Find values of m from the equation φn (m) = 0 ⇒ a0 + a1 m + a2 m2 + · · · + an mn = 0.
Let the solutions are m1 , m2 , . . . mk .
CASE-I: If φn (m) has no repeated root, then φ′n (m) 6= 0.
Hence, in that case, ci = − φφn−1 (mi )
′ (m )
i
=⇒ ci φ′n (mi ) + φn−1 (mi ) = 0.
n
Then the asymptotes are y = mi x + ci .
CASE-II: If φ′n (mi ) = 0, that is, φn (m) has a repeated root at mi , and if φn−1 (mi ) 6= 0,
then there is no asymptote corresponding to the value mi .
CASE-III: If φ′n (mi ) = φn−1 (mi ) = 0, for some value of mi , then the value of ci is
9
determined from:
c2 ′′ c
φn (mi ) + φ′n−1 (mi ) + φn−2 (mi ) = 0.
2! 1!
Similarly, if φ′′n (mi ) = φ′n−1 (mi ) = φn−2 (mi ) = 0, then the value of c is determined
from:
c3 ′′′ c2 c
φn (mi ) + φ′′n−1 (mi ) + φ′n−2 (mi ) + φn−3 (mi ) = 0.
3! 2! 1!
Example: Find the asymptotes of the curve
c h
g = 0. So m = 2, m = 1, m = −1
Now φ (m) = 0 =⇒ 2 − m − 2m +am 2 3
=− B
3 1 2 3
For m = 2, c = −
1 1
S
φ2 (m2 )
φ′3 (m2 )
= −4.−4+8.2
3.22 −4.2−1
10
Problem Set
3. Find the asymptotes of the curve 3x3 −2x2 y−4xy 2 +y 3 +3x2 +4xy−7x+10 = 0.
7 )
9. Verify that r cos θ ± b = a are asymptotes of the curve r = a sec θ + b tan θ.
1
20
10. Find the asymptotes of the curve (2r − 3) sin θ = 5.
(
11. Find the asymptotes of the curve y (x i − a ) = x (x − 4a ).
ch
2 2 2 2 2 2
Ans: x = ±a, y = ±x
a g
12. Find the asymptotes of the B 3 2 2
curve y − xy − x y + x + x − y − 1 = 0. 3 2 2
S
Ans: y + x = 0, y = x, y = x + 1
15. Find the asymptotes of the curve y(y − x)2 (y − 2x) + 3x2 (y − x) − 2x2 = 0.
3x 1
16. Find the asymptotes of the curve y = 2
log(e − 3x
).
17. Find the asymptotes of the curve y = a log sec( xa ). Ans: x = (2nπ ± π2 )a.
3x 1
Problem No 16: Find the asymptotes of the curve y = 2
log(e − 3x
).
1
Solution: The function is defined and continuous when x < 0 and when x − 3e
> 0.
Since the function is continuous at every point of the domain of definition, vertical
1
asymptotes can exist only at the end points x = 0 and x = 3e .
Vertical asymptotes
3x 1 1 log(e + t) 1
(i) lim y = lim log(e − ) = − lim = 0. [when t = − ]
x→0− x→0− 2 3x 2 t→+∞ t 3x
11
∴ x = 0 is not a vertical asymptote.
(ii)
3 1
lim y= lim x log(e − ) = −∞.
1
x→ 3e + 2 x→ 3e +
1 3x
1
∴ the line x = 3e
is a vertical asymptote.
Oblique asymptote
y 3 1 3
m = lim = lim log(e − ) = .
|x|→∞ x 2 |x|→∞ 3x 2
and
1
3x 1 3x 3 log(1 − 3xe
) 1
c = lim (y − mx) = lim { log(e − ) − } = lim { 1 }=− .
|x|→∞ |x|→∞ 2 3x 2 2 |x|→∞ x
2e
∴ the line y = 32 x − 1
2e
is an oblique asymptote of this curve.
7 )
Thus the curve has a vertical asymptote x =
0
1
3e 1
and an oblique asymptote y = 32 x− 2e
1
.
( 2
Problem No 17: Find the asymptotes of the icurve y = a log sec( ).
chnot parallel to y-axis.
x
a
∴ m = lim|x|→∞ y
S
= lim|x|→∞
a log sec( x
a
)
(∞
|x|→∞ x
form) = lim|x|→∞
d
dx
a log sec( x
d
a
)
x x ∞ x
dx
= lim|x|→∞ tan( xa ) = does not exist.
∴ there is no oblique asymptotes.
To find, if there be any asymptote parallel to y-axis, when |y| → ∞, then |a log sec( xa )| →
∞ or sec( xa ) → ∞ or xa → 2nπ ± π2 , n = 0, ±1, ±2, . . . .
x
∴ the required asymptotes are a
= 2nπ ± π2 , or x = (2nπ ± π2 )a, where n =
0, ±1, ±2, . . . .
To find, if there be any asymptotes parallel to x-axis.
y
We have, e a = sec( xa ). From this relation we see that there is no finite y such that
x → ∞.
Therefore, there is no asymptotes parallel to x-axis.
Problem No 15: (Using different method) Find the asymptotes of the curve y(y −
x)2 (y − 2x) + 3x2 (y − x) − 2x2 = 0.
Solution: The possible asymptotes are (i) one parallel to y = 0, (ii) two parallel to
y − x = 0, (iii) one parallel to y − 2x = 0.
Asymptote Parallel to y = 0, we have y(y − x)2 (y − 2x) + 3x2 (y − x) − 2x2 = 0.
12
Equating the coefficients of highest available power of x to 0, we get −(2y + 3) = 0
or 2y = −3.
Asymptote Parallel to y − x = 0,
3x2 −2x2
(y − x)2 + lim + lim = 0.
y=x
x→∞
y(y − 2x) x→∞
y=x y(y − 2x)
or (y − x − 2)(y − x − 1) = 0 or y − x − 2 = 0 and y − x − 1 = 0.
Asymptote Parallel to y − 2x = 0,
3x2 −2x2
(y − 2x) + lim + lim = 0.
y=2x y(y − x) y=2x y(y − x)2
x→∞ x→∞
or 2y − 4x + 3 = 0.
Hence the asymptotes are 2y = −3, y − x − 2 = 0, y − x − 1 = 0 and 2y − 4x + 3 = 0.
*******
7 )
1
( 20
i
Concavity, Convexity and Point of Inflection
g ch
a
Definition We call a function f (x) convex on interval I if, for all a, b ∈ I, a < b, and
SB
each x ∈ (a, b),
f (b) − f (a)
f (x) < (x − a) + f (a).
b−a
Definition We call a function f (x) concave on interval I if, for all a, b ∈ I, a < b,
and each x ∈ (a, b),
f (b) − f (a)
f (x) > (x − a) + f (a).
b−a
Definition We say that A = [a, f (a)] is an inflection point of the graph of f , if
• on one of the intervals (a − δ, a), (a, a + δ) the function f is convex and on the
other of them it is concave.
13
• if f ′′ (x) < 0 at each x ∈ (a, b), then f is concave on (a, b);
• if f ′′ (c) = 0 for some c ∈ (a, b), then f is Point of inflection at c: concavity
switches to convexity (or vice versa).
Example: Identify convexity, concavity intervals and inflection points of the function
x3 x2 1
f (x) = − − 2x + .
3 2 3
Problem Set
7 )
0 1 points of the function f (x) =
2
1. Find convexity, concavity intervals and inflection
(
x − sin x.
i
2. Show that the function f is convexcifhand only if the function −f is concave.
a g
S B
3. Determine the concavity/convexity 1 2
of f (x) = − x + 8x − 3.
3
3 2
4. Find numbers a and b such that the graph of the function f (x) = ax + bx
passes through (−1, 1) and has an inflection point at x = 21 . Ans. a = − 25 ,
b = 53
5. Find a, b and c such that the graph of the function f (x) = ax3 + bx2 + cx + 2
passes through (−1, 1) and (2, 2) and has an inflection point at x = 1. Ans.
1 1 1
a = 6, b = −2, c = 3
*******
lim f (x, y) = l.
(x,y)→(a,b)
14
Definition 2: A function f (x, y) is said to tend to the limit l as x → a and y → b
if and only if corresponding to the positive number ǫ therep exists another positive
number δ = δ(ǫ) such that |f (x, y) − l| < ǫ, whenever 0 < (x − a)2 + (y − b)2 < δ.
Definition 3: If a function f (x, y) has distinct limits as (x, y) approaches a point
(a, b) along two distinct paths, the limit lim(x,y)→(a,b) f (x, y) does not exist.
Definition 4: A function f (x, y) is said to be continuous at the point (a, b) if (i)
lim(x,y)→(a,b) f (x, y) exists and is equal to l, say, and (ii) f (a, b) = l.
Problem Set
1. Show that
x2 − y 2
lim xy = 0.
(x,y)→(0,0) x2 + y 2
2 2
5. Given that f (x, y) = xx2 +y
−y 2 2
2 , where, x + y 6= 0, is it possible to assign a value
√ xy
(
if (x, y) 6= (0, 0),
f (x, y) = x2 +y 2
0 if (x, y) = (0, 0)
ǫ2 ǫ2
∴ | √ xy
p
2
− 0| = |r cos θ sin θ| ≤ r = x2 + y 2 < ǫ, if x2 < 2
and y 2 < 2
.
x +y 2
15
Thus | √ xy
p
− 0| < ǫ whenever 0 < x2 + y 2 < ǫ.
2 x +y 2
xy
∴ lim p = 0.
(x,y)→(0,0) x2 + y 2
Hence, f (x, y) is continuous at (0, 0).
Example: Show that the function f (x, y) defined by
( 2 2
xy xx2 −y 2 when (x, y) 6= (0, 0),
f (x, y) = +y
0 when (x, y) = (0, 0)
(
ǫ = δ isuch that
ch+ y < δ.
2
So for any δ > 0 there exists an p
|xyx2 −y 2
− 0| < ǫ whenever 0 < x
x2 +y 2
a g 2 2
B
S ∴ lim f (x, y) = 0.
(x,y)→(0,0)
*******
Partial Differentiations
Problem Set
∂u ∂u x2 −y 2
1. Find ∂x
and ∂y
when u = tan−1 x2 +y 2
.
√ ∂2U ∂2U
2. If U = xy, find the value of ∂x2
+ ∂y 2
.
16
∂2z ∂ z 2
6. If z = f (x + ay) + φ(x − ay) show that ∂y 2
= a2 ∂x 2.
2 2 2
10. If u = xφ( xy ) + yψ( xy ), prove that x2 ∂∂xu2 + 2xy ∂x∂y
∂ u
+ y 2 ∂∂yu2 = 0.
∂2u ∂ u 2 ∂2u
11. If u = x φ(x + y) + y ψ(x + y) prove that ∂x2
− 2 ∂x∂y + ∂y 2
= 0.
3 3
12. If f (x, y) = tan−1 xx−y+y
, using Euler’s theorem, show that x2 fxx + 2xyfxy +
y 2 fyy = (2 cos 2f − 1) sin 2f.
x3 +y 3
13. If u = tan−1 x−y
, prove that xux + yuy = sin 2u.
1
15. If u = (x2 + y 2 + z 2 ) 2 , prove that uxx + uyy + uzz = u2 .
( 20
i
gch
∂z ∂z
16. If z = u3 + v 3 , where u = sin xy and v = y 2 , find ∂x
and ∂y
.
17. If ∂u = ∂y∂v
and ∂u ∂v
= − ∂x
B a
, where u and v are functions of x and y, prove
S
∂x ∂y
that ∂r = r ∂θ and r ∂θ = − ∂v
∂u 1 ∂v 1 ∂u
∂r
, where x = r cos θ, y = r sin θ. [Hint: ∂u
∂r
=
∂u ∂x ∂u ∂y
∂x ∂r
+ ∂y ∂r .]
*******
Jacobian
Problem Set
∂(x,y)
1. If x = r cos θ, y = r sin θ, find ∂(r,θ)
.
∂(x,y)
2. If x = a cos θ cosh φ, y = a sin θ sinh φ, find ∂(φ,θ)
= 12 a2 (cosh 2φ − cos 2θ).
∂(u,v) 1 y 2 −x2
3. If u3 + v 3 = x + y , u2 + v 2 = x3 + y 3 , show that ∂(x,y)
= 2 uv(u−v)
.
∂(x,y,z)
4. If x = r sin θ cos φ, y = r sin θ sin φ, z = r cos θ, show that ∂(r,θ,φ)
= r2 sin θ.
∂(x,y,z)
5. If x = r + s + t, y = st + tr + rs, z = rst, show that ∂(r,s,t)
= (r − s)(r − t)(s − t).
∂(r,s,t)
Also, find the value of ∂(x,y,z)
.
17
∂(x,y)
6. If x + y = u and x = uv then show that ∂(u,v)
= −u. Also, find the value of
∂(u,v)
∂(x,y)
.
1 1
7. If u = x(1 − r2 )− 2 and v = y(1 − r2 )− 2 where r2 = x2 + y 2 , find the value of
∂(u,v)
∂(x,y)
.
7 )
1
11. If u = x + 2y + z, v = x − 2y + 3z, and w = 2xy − xz + 4yz − 2z 2 , then find the
value of ∂(u,v,w)
∂(x,y,z)
( 20
. Is there relation between u, v, and w? If yes, then what is the
relation between them?
h i
g c
a *******
SB
SEQUENCE AND SERIES
2. Comparison Test
18
P P P
(a) If two positive term series un and P vn be such that (i) vn converges,
(ii) un ≤ vn for all values of n, then un also converges.
P P P
(b) If two positive term series un and P vn be such that (i) vn diverges,
(ii) un ≥ vn for all values of n, then un also diverges.
P P
(c) If two positive term series un and vn be such that
un
lim = finite (6= 0),
n→∞ vn
P P
then un and vn both converge or diverge.
P
3. D’Alembert’s Ratio Test In positive term series un , if
un+1
lim = λ,
n→∞ un
then the series converges for λ < 1 and diverges for λ > 1. (for λ = 1 there is
no conclusion.)
4. Cauchy’s Root Test In positive term series
P
un , if
7 )
1
20
1
n→∞
i (
lim unn = λ,
ch
then the series converges for λ < 1 and diverges for λ > 1. (for λ = 1 there is
g
no conclusion.)
a
5. P
SB
Cauchy Condensation Test Suppose a1 ≥ a2 ≥ a3 ≥ a4 ≥ · · · ≥ 0. Then
∞
n=1 an converges if and only if
∞
X
2k a2k = a1 + 2a2 + 4a4 + 8a8 + . . .
n=0
converges.
6. Leibnitz’s Rule An alternating series u1 − u2 + u3 − u4 + . . . . converges if (i)
each term is numerically less than its preceding term, and (ii) limn→∞ un = 0.
(limn→∞ un 6= 0, the given series is oscillatory.)
Problem Set
∞ ∞ 1 ∞ ∞
X n2 X sin n
X sin n1 X 1
(vi) (vii) (viii) √ (ix)
n=1
3n n=1
n n=1
n n=2
n(log n)p
(where p > 0).
19
2. Investigate the convergence of the series:
12 12 .32 12 .32 .52 2
(a) 22
+ 22 .42
x + 22 .42 .62
x + . . . ∞. (x 6= 0) Ans: ?
x2 x3 x4
(b) x + 2
+ 3
+ 4
+ . . . , x ≥ 0. Ans: ?
(
n] is non-absolutely convergent series. Can you give some other examples?
i
7. Show that ∞ 1
ch
P
n=1 n! is convergent.
g
a+ (2 + 2a) + (2 + 3a) + . . . diverges for any real
values of a.
S B
8. Prove that the series 2 + (2 + a)
P∞ an+1
9. Consider
P∞ a series of positive term as n=1 an . If limn→∞ an = 1, can you say
that n=1 an is convergent? Justify your answer.
1
10. Show that n≥2 (−1)n n log
P
n
is non-absolutely convergent series. Using the
Cauchy condensation test, determine the convergence of the following series:
P∞ 1
(a) n=2 n log n . Ans: Divergent.
P∞ 1
(b) n=2 (n log n)2 . Ans: Convergent.
P∞ 1
(c) n=2 (n log n)p . Ans: Conv if p > 1, div if p ≤ 1.
P∞ 1
(d) n=2 (n log n)(log(log n)) . Ans: Divergent.
*******
1. Define order and degree of an ODE. Find the order and degree of the ODEs
20
dy 23 2
(a) y + ( dx ) = 1 + x and ( ddxy )2 + y = dy
dx
.
2. Homogeneous equation: If a function f (x, y) can be expressed in the form
xn φ( xy ) or in the form y n ψ( xy ), then f (x, y) is said to be a homogeneous function
of degree n.
3. Solve the following differential equations (Put y = vx):
x 3
(a) x2 y dx − (x3 + y 3 ) dy = 0 Ans: − 3y 3 + log y = c
2 2
(d) y(2xy + 1) dx + x(1 + 2xy + x y ) dy = 0 (xy = v).
dy
Note: The above equations can be written in the form dx
= f (x, y), where
f (x, y) is a homogeneous function.
4. Rules for finding integrating factors:
7 )
(a) If M x + N y 6= 0 and the equation be homogeneous, then M x+N 1
y
is an
1
integrating factor (I.F.) of the equation M dx + N dy = 0.
( 20
(b) If M x − N y 6= 0 and the equation can be written as {f (xy)}ydx +
1
i
ch
{F (xy)}xdy = 0, then M x−N y
is an integrating factor of the equation
M dx + N dy = 0.
(c) If N1 ( ∂M ∂N
a g R
f (x)dx
SB
∂y
− ∂x
) be a function of x alone, say f (x), then e is an
integrating factor of the equation M dx + N dy = 0.
R
(d) If M1 ( ∂N
∂x
− ∂M
∂y
) be a function of y alone, say φ(y), then e φ(y)dy
is an
integrating factor of the equation M dx + N dy = 0.
5. Solve the following differential equations:
(i) x dx + y dy + x dy−y dx
x2 +y 2
=0 Ans: x2 + y 2 + 2 tan−1 ( xy ) = C
(ii) (x y − 2xy ) dx + (3x y − x3 ) dy = 0
2 2 2
Ans: xy + 3 log xy + log x = c
dy (xy sin xy+cos xy)y
(iii) dx
= (cos xy−xy sin xy)x
Ans:
(iv) (3x2 y 4 + 2xy) dx + (2x3 y 3 − x2 ) dy = 0 Ans:
dy
(v) x2 dx +y =1 Ans:
dy dy
(vi) y + x2 dx
2
= xy dx Ans:
dy 2 x
(vii) dx = y e x+2xy
2 Ans: I.F. = y −2
(viii) (y 2 + x2 + 2x) dx + 2y dy = 0 Ans: I.F. = ex (using 4c)
(ix) (x2 + y 2 + 2x) dx + 2y dy = 0 Ans:
(x) (x4 + y 4 ) dx − xy 3 dy = 0. Ans:
6. Solution of first order but not of first degree ODE:
Start with a first order and n-th degree ODE:
pn + P1 pn−1 + ... + Pn−1 p + Pn = 0,
dy
where Pi are functions of x and y; p ≡ dx
.
Here we mainly discuss the following methods for solving.
21
(a) Solvable for p: If it can be expressed as
{p − f1 }{p − f2 }....{p − fn } = 0.
φ(x, p, c) = 0. (2)
Next we eliminate p from (1) and (2), we get the required solution. And
7 )
(c) Solvable for x: If it can be expressed as
1
( 20
i
x = f (y, p). (3)
g ch
Differentiating both sides of (3) with respect to y, we get an equation of
a
SB
the form
1 dp
= F (y, p, ).
p dy
It can be solved to get a solution of the form
φ(y, p, c) = 0. (4)
Next we eliminate p from (3) and (4), we get the required solution.
8. Clairut′ s Equation
A differential equation of the form y = px+f (p) is known as Clairut′ s Equation.
A Clairaut’s equation is a differential equation of the form
dy dy
y(x) = x + f .
dx dx
22
To solve such an equation, we differentiate with respect to x, yielding
d2 y
2
dy dy ′ dy d y
= +x 2 +f ,
dx dx dx dx dx2
dy
d2 y
so 0 = x + f ′ dx dx2
.
d2 y dy
or 0 = x + f ′
Hence, either 0 = dx2 dx
.
dy
In the former case, C = dx for some constant C. Substituting this into the
Clairaut’s equation, we have the family of straight line functions given by y(x) =
Cx + f (C), the so-called general solution of Clairaut’s equation.
dy
The latter case, 0 = x + f ′ dx
, defines only one solution y(x), the so-called
singular solution, whose graph is the envelope of the graphs of the general so-
lutions. The singular solution is usually represented using parametric notation,
dy
as (x(p), y(p)), where p represents dx .
9. Solve the following differential equations and find the singular solutions:
7 )
√
1
20
p
(a) y = px + 1 + p2 Ans: y = cx + 1 + c2 ; x2 + y 2 = 1.
(b) y = px + a
i ( Ans: y = cx + ac ; y 2 = 4ax.
ch
p
√
g
2 2
Ans: y = cx + a2 c2 + b2 ; xa2 + yb2 = 1.
p
(c) y = px + a2 p 2 + b 2
a
SB
(d) y = px + p − p2 Ans: y = cx + c − c2 ; 4y = (x + 1)2 .
12. Solution Method for linear differential equation of the form f (D)y = 0:
If the roots of the auxiliary equation f(m)=0 are m1 , m2 and m3 , then the
solutions of the given ODE are
23
(c) If m1 = m2 = m3 , y = (C1 + C2 x + C3 x2 )em1 x ,
(d) If m1 = α+iβ, m2 = α−iβ and m3 , y = eαx (C1 cos βx+C2 sin βx)+C3 em3 x .
13. Particular Integral (P.I.) By Short Method (Rules) of the form f (D)y =
X(x):
1
(a) If X = eax , P.I. = f (a) eax , provided f (a) 6= 0,
r
x e ax
= r!g(a) , where f (D) = (D − a)r g(D), g(a) 6= 0.
(b) If X = sin ax/ cos ax,
1
P.I. = φ(D1 2 ) sin ax = φ(−a 2
2 ) sin ax, provided φ(−a ) 6= 0, where f (D) =
φ(D2 ).
If φ(−a2 ) = 0 then P.I. = D21−a2 sin ax = − 2a x
cos ax and D21−a2 cos ax =
x
2a
sin ax.
1
(c) If X = V eax , V being functions of x only, P.I. = eax f (a+D) V.
1 1
(d) If X = xV , V being functions of x only, P.I. = x f (D) V − f ′ (D) [f (D)]2V .
7 )
(e) If X = P (x), where P (x) is a Polynomial function of x, then we have
1
20
to expand [f (D)]−1 and arrange the terms in ascending powers of D and
operate with P (x).
i (
h
Remark: With the help of Bernstein coefficient nr = n.(n−1).(n−2)...(n−r+1)
,
(1 + x)n = 1 + x+ x2 + x3 + . . . .
1 2 3
24
x sin 3x
(k) (D2 + 9)y = sec 3x P.I. = 3
− 19 cos 3x log sec 3x
1
(l) (D3 − 3D2 + 4D − 2)y = ex + cos x P.I. = xe + x
10
(cos x + 3 sin x)
x2
(m) (D2 − 1)y = x sin x + xex P.I. = 4
ex − x2 sin x − cos x
2
(n) (D2 + 2D + 2)y = x2 e−x P.I. = e−x [x2 − 2]
1 3x
(o) (D2 − 9)y = e3x cos x P.I. = − 37 e [cos x − 6 sin x]
ex x 4
(p) (D2 − 2D + 1)y = x2 ex . P.I. = 12
(q) (D2 + 1)y = sin x sin 2x P.I. = 14 x sin x + 1
16
cos 3x
2 1
(r) (D − D − 2)y = sin 2x P.I. = 20
[cos 2x − 3 sin 2x]
2 3 3
(s) (D − 1)y = x + 2x + 1 P.I. = x + 4x + 1
7 )
1
c1 f (t) + c2 g(t) = 0; ∀t ∈ I.
0
( 2
If we must choose c = 0 = c , then we
1 2
Since we are considering only two c
i say f and g are linearly independent.
h linear dependence is equivalent in
functions,
this special case to one function a g
being a scalar multiple of the other.
S
If two differentiable functionsBf and g are linearly dependent, then their Wron-
skian is zero for all t ∈ I, i.e.,
25
independent in the interval (−2, 2) but surprisingly Wronskian is not zero for
all x ∈ I:
W [f ; g] = 6x2 6= 0
as long as x 6= 0.
S B
−x 2
(h) Can you form a second order linear homogeneous ODE whose two inde-
pendent solutions are f (t) = 2t and g(t) = 3t2 in (−2, 2)?
be the complementary function of the equation (5), where A and B are con-
stants and u and v are functions of x and are independent solutions of the
corresponding homogeneous equation. Thus
d2 u du d2 v dv
+ P + Qu = 0 and + P + Qv = 0. (7)
dx2 dx dx2 dx
Let us assume
y = φu + ψv (8)
26
is complete primitive of (5), where we take φ and ψ in place of A and B and
they are no longer constants but functions of x, to be so chosen that (8) will
satisfy (5).
Differentiating (8) with respect to x, we get
dy du dv dφ dψ
=φ +ψ +u +v .
dx dx dx dx dx
dφ dψ
u +v = 0, (9)
dx dx
so that
dy du dv
=φ +ψ
dx dx dx
Differentiating this once again with respect to x, we get
7 )
1
d2 y d2 u d2 v dφ du dψ dv
( 20
dx2
=φ 2 +ψ 2 +
dx
h dx
+
i
dx dx dx dx
.
g c
Substituting these values of y,a and dy d2 y
in (5) and after rearrange, we get
SB
dx dx2
d2 u du d2 v dv dφ du dψ dv
φ( 2 + P + Qu) + ψ( 2 + P + Qv) + + =X
dx dy dx dy dx dx dx dx
dφ du dψ dv
+ =X (10)
dx dx dx dx
by virtue of (7).
dφ dψ
Solving for dx
and dx
from (9) and (10), we get
dφ vX dψ uX
= du dv
and = du dv
dx v dx − u dx dx v dx − u dx
Integrating we get
vXdx uXdx
Z Z
φ = C1 + du dv
and ψ = C2 +
v dx − u dx v du dv
− u dx
dx
Substituting these values of φ and ψ in (8), we get the complete solution of the
equation (5), C1 and C2 being arbitrary constants.
27
d2 y
(a) dx2
+ a2 y = sec ax Ans:
y = C1 cos ax + C2 sin ax + xa sin ax + 1
a2
cos ax log(cos ax).
d2 y
(b) dx2
+ 4y = 4 tan 2x Ans:
y = C1 cos 2x + C2 sin 2x − cos 2x log(sec 2x + tan 2x).
d2 y 2 x
(c) dx2
−y = 1+ex
Ans: y = C1 ex + C2 e−x + ex log 1+e
ex
− ex log(1 + ex ) − 1.
d2 y
(d) dx2
+ y = cosec x Ans: y = C1 cos x + C2 sin x − x cos x + sin x log sin x.
d2 y
(e) dx2
+ y = cos x Ans: y = C1 sin x + C2 cos x + 12 x sin x + 14 cos x.
d2 y
(f) dx2
+ 4y = sin 2x Ans: y = C1 cos 2x + C2 sin 2x − 14 x cos 2x.
d2 y dy e−x
(g) dx2
+ 2 dx +y = x2
Ans: y = (C1 + C2 x)e−x − e−x (1 + log x).
d2 y
(h) dx2
+y =x Ans: y = C1 cos x + C2 sin x + x.
7)
(a) (D − 7)x + y = 0, (D − 5)y − 2x = 0; where D ≡ dt
(b) (D + 5)x + y = et , (D + 3)y − x = e2t ;
(c) (D − 3)x − 4y = 0, (D + 1)y + x =2 0 1
2
dy
S (x, y, z), Q ≡ Q(x, y, z), R ≡ R(x, y, z):
dx dz
(a) xy
= y2
= xyz−2x2
Ans: x = C1 y, x = log(yz − 2x) − log y + C2 .
adx bdy cdz
(b) = xz(c−a)
yz(b−c)
= xy(a−b) Ans: ax2 + by 2 + cz 2 = C1 ,
a x + b2 y 2 + c 2 z 2 = C 2 .
2 2
xdx dy dz
(c) z 2 −2yz−y 2
= y+z
= y−z
Ans: x2 + y 2 + z 2 = C1 , x2 − z 2 − 2yz = C2 .
dx dy dz
(d) 1+y
= 1+x
= z
Ans: z(x − y) = C1 , z = C2 (x + y + 2).
dx dy dz 1
(e) y2
= x2
= x2 y 2 z 4
Ans: x3 − y 3 = C1 , y 3 + z3
= C2 .
dx dy dz
(f) yz
= zx
= xy
Ans: x2 − y 2 = C1 , x2 − z 2 = C2 .
dx dy dz
(g) z
= −z
= z 2 +(x+y)2
Ans: x + y = C1 , log[(x + y)2 + z 2 ] − 2x = C2 .
dx dy dz
21. Geometrical interpretation of the equations P
= Q
= R
.
We know, from the geometry of three dimensions, that the direction cosines of
the tangent to a curve are dx , dy , dz , that is are in the ratio dx : dy : dz.
ds ds ds
Hence geometrically these equations represent a system of curves in space, such
that the direction cosines of the tangent to it at any point (x, y, z) are propor-
tional to P , Q, R.
28
USEFUL FORMULAS
x −x ex +e−x
1. cosh x = e −e
2
, sinh x = 2
d
, dx (cosh x) = sinh x, d
dx
(sinh x) = cosh x,
2 2
cosh x − sinh x = 1.
2. ∞
x
X xn 1 n x n
e = , lim (1 + ) = e and lim (1 + ) = ex .
n=0
n! n→∞ n n→∞ n
3. The equation of the cardiod is r = a(1 + cos θ) and shape of the equation is in
Figure1.
7 )
1
( 20
h i
g c
a
Figure 1: Cardiod
SB *******
BOOKS: (TEXT/REFERENCES)
29