Exercises Power Series
Exercises Power Series
Problem 1 (10 pts) Find the radius of convergence and interval of convergence of the series
X∞ n(x + 2)n
.
n=1 5n−1
n(x + 2)n
Solution: Do the ratio test for the absolute convergence. Let an = . Then,
5n−1
¯ (n + 1)(x + 2)n+1 n−1
¯ ¯ ¯ ¯
¯ a n+1 ¯ 5 ¯
¯ a ¯ = ¯ ·
¯ ¯ ¯ ¯
5 n n(x + 2) ¯n
n
n +1 |x + 2|
= |x + 2| −−−−→ .
5n n→∞ 5
|x + 2|
Hence, the power series converges absolutely if < 1. From this, we can see that the
5
radius of convergence is R = 5 . Also, the interval of convergence is −5 < x + 2 < 5, i.e.,
−7 < x < 3.
Let’s check the convergence when x is at the boundary points. For x = −7, the series be-
comes: ∞ n ∞
X n(−5)
5n(−1)n .
X
n−1
=
n=1 5 n=1
Since lim 5n(−1)n 6= 0, this series does not converge (the n th Term Test for Divergence).
n→∞
So, we cannot include x = −7 in the interval of convergence. How about x = 3? This leads
to ∞ n ∞
X n5 X
n−1
= 5n.
n=1 5 n=1
Clearly this diverges (again via the n th Term Test for Divergence). So, we cannot include
x = 3 in the interval of convergence either. Hence the interval of convergence is: (−7, 3) or
−7 < x < 3 .
f (x) = ln x
= P 2 (x) + R 2 (x)
f 00 (2) f 000 (c)
= f (2) + f 0 (2)(x − 2) + (x − 2)2 + (x − 2)3 2 < c < x or x < c < 2
2 3!
1 1 1
= ln 2 + (x − 2) − (x − 2)2 + 3 (x − 2)3 .
2 8 3c
1 1
Hence, we have P 2 (x) = ln 2 + (x − 2) − (x − 2)2 .
2 8
Solution to (b): The approximation error is:
1 1 1
|R 2 (x)| = 3
|x − 2|3 ≤ 3 < ,
3c 3c 3
since |x − 2| ≤ 1 and 1 < c < 3. Therefore, the accuracy of approximation by P 2 (x) for
1
1 ≤ x ≤ 3 is bounded by ≈ 0.333 , which is the worst-case scenario.
3
Problem 3 (10 pts) State and prove the Cauchy-Schwarz Inequality. Note that you also need to
state and prove the condition for the equality to hold.
|u · v| ≤ |u| |v| where the equality holds if and only if u and v are parallel to each other.
The proof of this inequality is based on the definition of the dot product. In other words, let
θ be the angle between u and v. Then,
From the above derivation, it is also clear that the equality holds if and only if cos θ = 0
(which includes the case where either u or v is the zero vector), which is the same as saying
that u and v are parallel to each other.
Problem 4 (10 pts) Let P (1, 4, 6), Q(−2, 5, −1), R(1, −1, 1).
1 ¯¯−−→ −→¯¯ 1 1p 1p 5p
¯PQ × P R ¯ = |〈−40, −15, 15〉| = 2050 = 25 · 82 = 82 .
2 2 2 2 2
Solution to (b): You can compute the distance d from P to the line QR by the formula based
on the cross product, i.e.,
¯−−→ −−→¯
¯ ¯
¯QP × QR ¯
d= ¯−−→¯ .
¯QR ¯
¯ ¯
since both amounts to the area of the same parallelogram, which is the twice as the area
−−→
of the triangle 4PQR . Since QR = 〈3, −6, 2〉, we have
p
5 82 5p
d=p = 82 .
32 + (−6)2 + 22 7
Alternatively, you can reason as follows: the distance d can be computed by the fol-
lowing basic formula
1 ¯−−→¯
Area of 4PQR = d · ¯QR ¯ .
¯ ¯
2
−−→
Since QR = 〈3, −6, 2〉, we have
p
2 × 52 82 5p
d=p = 82 .
32 + (−6)2 + 22 7
Problem 5 (10 pts) Find the limit if it exists, or show that the limit does not exist.
(a) (5 pts)
y4
lim .
(x,y)→(0,0) x 4 + 3y 4
y4 m4x4 m4 m4
lim = lim = lim = .
(x,y)→(0,0) x 4 + 3y 4 x→0 x 4 + 3m 4 x 4 x→0 1 + 3m 4 1 + 3m 4
This value depends on m . Since the limit depends on how the point (x, y) approaches
(0, 0), the limit does not exist .
Solution to (b): Consider the polar coordinates: x = r cos θ , y = r sin θ , where r ≥ 0, 0 ≤
θ < 2π.
This is true regardless of the direction of approach θ . Hence, the limit exists and its value is 0 .
where k is an arbitrary positive integer is a solution of the heat conduction equation over the
interval x ∈ [0, π]
u t = α2 u xx ,
with the boundary condition
∂u 2 2 ∂2 u 2 2
ux = = e−α k t · k cos kx, u xx = = e−α k t · (−k 2 ) sin kx.
∂x ∂x 2
Hence,
2 2
k t
α2 u xx = −α2 k 2 e−α sin kx = u t .
Finally, we can easily check that the boundary condition is satisfied since
2 2 2 2
k t k t
u(0, t ) = e−α sin k0 = 0, u(π, t ) = e−α sin kπ = 0,
for any t ≥ 0.
f (x, y, z) = ln(x y 2 z 3 ) = ln x + 2 ln y + 3 ln z.
So, ∇ f (1, 2, 3) = 〈1, 1, 1〉. Now, f increases most rapidly in the direction of ∇ f . Hence,
this direction is 〈1, 1, 1〉 .
(It is up to you
¿ À whether you want to make it to the normal vector of length 1, i.e.,
1 1 1
p , p , p , but it is not necessary in this case.)
3 3 3
Solution to (b): f decreases most rapidly in the direction of −∇ f . Hence, it is −〈1, 1, 1〉 =
〈−1, −1, −1〉 .
Solution to (c): f does not change in the direction orthogonal to ∇ f . Let u = 〈u 1 , u 2 , u 3 〉 be
this directional vector. Then,
u · ∇ f (1, 2, 3) = 0 ⇐⇒ u = 〈u 1 , u 2 , u 3 〉 · 〈1, 1, 1〉 = 0 ⇐⇒ u 1 + u 2 + u 3 = 0.
Since we have
2x 2y 2z
F x (x, y, z) = , F x (x, y, z) = , F x (x, y, z) = ,
a2 b2 c2
the equation of the tangent plane becomes
2x 0 2y 0 2z 0 x 0 (x − x 0 ) y 0 (y − y 0 ) z 0 (z − z 0 )
(x − x 0 ) + (y − y 0 ) + (z − z 0 ) = 0 ⇐⇒ + + =0
a2 b2 c2 a2 b2 c2
x 0 x y 0 y z 0 z x 02 y 02 z 02
⇐⇒ + 2 + 2 = 2 + 2 + 2 = 1.
a2 b c a b c
In the last line, the righthand side becomes 1 since the point P 0 (x 0 , y 0 , z 0 ) is on this
ellipsoid.
2x 0 2y 0 2z 0
Solution to (b): At the point P 0 (x 0 , y 0 , z 0 ), the gradient vector is ∇F (x 0 , y 0 , z 0 ) = 〈 2 , 2 , 2 〉.
a b c
The normal line passes through the point P 0 is parallel to this gradient vector. Hence,
the equation of this normal line is
³ ´
2x 0 2
x = x 0 + t · a2
= x 0 1 + t
a2 ´
³
2y 0
y = y 0 + t · b 2 = y 0 1 + b22 t ,
³ ´
z = z 0 + t · 2z20 = z 0 1 + 22 t
c c
p
Problem 10 (10 pts) Find the linearization of f (x, y, z) = x 2 + y 2 + z 2 at the point (3, 2, 6). Then
use it to approximate f (3.1, 1.9, 6.2) by the three decimal point number. Note that the true
value of f (3.1, 1.9, 6.2) is 7.188 in the three decimal point number.
We have
1 1 1
· (x 2 + y 2 + z 2 )− 2 · 2x = x(x 2 + y 2 + z 2 )− 2 .
f x (x, y, z) =
2
Since f (x, y, z) is symmetric with respect to x , y , z , we immediately have
1 1
f y (x, y, z) = y(x 2 + y 2 + z 2 )− 2 , f z (x, y, z) = z(x 2 + y 2 + z 2 )− 2 .
p p
Now, at the point (3, 2, 6), we have f (3, 2, 6) = 32 + 22 + 62 = 49 = 7. Hence,
3 2 6
L(x, y, z) = 7 + (x − 3) + (y − 2) + (z − 6) .
7 7 7
Problem 11 (10 pts) Find the absolute maximum and minimum values of
f (x, y) = x 2 − x y + y 2 + 1,
Ω = {(x, y) | x ≥ 0, y ≤ 2, y ≥ 2x}.
f x = 2x − y = 0, f y = −x + 2y = 0 ⇐⇒ y = 2x, x = 2y ⇐⇒ x = 4x ⇐⇒ x = 0.
Hence, x = y = 0 is the only critical point in Ω. Now we need to check whether this point is
local min., local max., or a saddle point. To do so, we compute
Since D(0, 0) > 0 and f xx (0, 0) = 2 > 0, the value f (0, 0) = 1 is the local minimum.
Step 3: Find the largest and the smallest values from the results of Steps 1 and 2.
Combining all the results we got so far, we can easily conclude that
the absolute maximum is f (0, 2) = 5 while the absolute minimum is f (0, 0) = 1 .
Problem 12 (10 pts) Use Lagrange multipliers to find the maximum and minimum values of the
function
f (x, y) = ex y subject to x 2 + y 2 = 1.
Note that we only consider the real values for x and y , not the complex values.
Solution: Let g (x, y) = x 2 + y 2 − 1 = 0 be the constraint. Let’s find the points (x, y) and the param-
eter λ satisfying
∇ f = λ∇g , g (x, y) = 0. (1)
We have
∇ f = 〈 f x , f y 〉 = 〈y ex y , x ex y 〉, ∇g = 〈g x , g y 〉 = 〈2x, 2y〉.
Hence, Eq. (1) leads to
y ex y = 2λx (2)
xy
xe = 2λy (3)
2 2
x +y = 1. (4)
We can see that λ 6= 0 because if λ = 0, then x = y = 0 from Eqn’s (2) and (3), which
contradicts Eq. (4). Similarly, x 6= 0 as well as y 6= 0. Hence, we can divide Eq. (2) by Eq. (3)
to get
y ex y 2λx y x
x y
= ⇐⇒ = ⇐⇒ x 2 = y 2 .
xe 2λy x y
1
This implies x = y or x = −y . Insert these into Eq. (4) gives us 2x 2 = 1 or x = ± p . So,
µ ¶ µ ¶ µ ¶ µ ¶ 2
1 1 1 1 1 1 1 1
we have (x, y) = p , p , p , − p , − p , p , or − p , − p . From Eq. (2), the
2 2 2 2 2 p 2 2 2 p
e 1 1 e
corresponding λ’s can be easily calculated as λ = , − p , − p , or , respectively.
2 2 e 2 e 2
Hence,
p
µ ¶ µ ¶
1 1 1 1
f p , p = f − p , − p = e,
2 2 2 2
µ ¶ µ ¶
1 1 1 1 1
f p ,−p = f −p , p = p .
2 2 2 2 e
Therefore, the maxima are
p
µ ¶ µ ¶
1 1 1 1
f p , p = f −p ,−p = e ,
2 2 2 2