MATH 2004 Homework Solution Han-Bom Moon

Homework 9 Model Solution

Section 14.7 ∼ 14.8.

14.7.6 Find the local maximum and minimum values and saddle points of f (x, y) =
xy − 2x − 2y − x2 − y 2 .

∇f = hfx , fy i = hy − 2 − 2x, x − 2 − 2yi

∇f = 0 ⇔ y − 2 − 2x = 0, x − 2 − 2y = 0
⇔ −2x + y = 2, x − 2y = 2 ⇔ x = −2, y = −2
Critical point: (x, y) = (−2, −2)

fxx = −2, fxy = 1, fyy = −2

At (x, y) = (0, 0),
−2 1
D= =3>0
1 −2
Also fxx = −2 < 0.
Therefore f (−2, −2) = 4 is a local maximum.

14.7.10 Find the local maximum and minimum values and saddle points of f (x, y) =
xy(1 − x − y).

f (x, y) = xy − x2 y − xy 2
∇f = hfx , fy i = hy − 2xy − y 2 , x − x2 − 2xyi
∇f = 0 ⇔ y − 2xy − y 2 = 0, x − x2 − 2xy = 0
⇔ y(1 − 2x − y) = 0, x(1 − x − 2y) = 0
From y(1 − 2x − y) = 0, y = 0 or y = 1 − 2x.
If y = 0, x(1 − x) = 0 ⇒ x = 0 or x = 1. So (0, 0), (1, 0) are critical points.
1
If y = 1 − 2x, x(1 − x − 2(1 − 2x)) = 0 ⇒ x(−1 + 3x) = 0 ⇒ x = 0 or x = .
3
1 1
x = 0 ⇒ y = 1, x = ⇒y= .
3 3
1 1
So (0, 1), ( , ) are the other critical points.
3 3

1
MATH 2004 Homework Solution Han-Bom Moon

In summary, there are four critical points

1 1
(0, 0), (1, 0), (0, 1), ( , ).
3 3
fxx = −2y, fxy = 1 − 2x − 2y, fyy = −2x

−2y 1 − 2x − 2y
D=
1 − 2x − 2y −2x

At (0, 0),
0 1
D= = −1 < 0.
1 0
So (0, 0) is a saddle point.
At (1, 0),
0 −1
D= = −1 < 0.
−1 −2
So (1, 0) is a saddle point.
At (0, 1),
−2 −1
D= = −1 < 0.
−1 0
So (0, 1) is a saddle point.
1 1
At ( , ),
3 3
− 23 − 13 1
D= = >0
− 13 − 23 3
1 1 2 1 1 1
Because fxx ( , ) = − < 0, f ( , ) = is a local maximum.
3 3 3 3 3 27
14.7.30 Find the absolute maximum and minimum values of f (x, y) = x + y − xy on D
where D is the closed triangular region with vertices (0, 0), (0, 2) and (4, 0).

∇f = hfx , fy i = h1 − y, 1 − xi
∇f = 0 ⇔ 1 − y = 0, 1 − x = 0 ⇔ x = 1, y = 1
Critical point: (1, 1).

fxx = 0, fxy = −1, fyy = 0

0 −1
D= = −1 < 0
−1 0
So (1, 1) is a saddle point.
The boundary consists of 3 line segments, 1) y = 0, 0 ≤ x ≤ 4, 2) x = 0, 0 ≤ y ≤
1
2, and 3) y = − x + 2, 0 ≤ x ≤ 4.
2

2
MATH 2004 Homework Solution Han-Bom Moon

On y = 0, 0 ≤ x ≤ 4,

f (x, y) = x ⇒ maximum = 4, minimum = 0.

On x = 0, 0 ≤ y ≤ 2,

f (x, y) = y ⇒ maximum = 2, minimum = 0.

1
On y = − x + 2, 0 ≤ x ≤ 4,
2
   
1 1 1 3
f (x, y) = f (x) = x + − x + 2 − x − x + 2 = x2 − x + 2.
2 2 2 2
 
d 1 2 3 3 3
x − x+2 =x− =0⇔x=
dx 2 2 2 2
3 7
f ( ) = , f (0) = 2, f (4) = 4.
2 8
7
So the maximum is 4, the minimum is .
8
Therefore, the absolute maximum is f (4, 0) = 4 and the absolute minimum is
f (0, 0) = 0.

14.7.31 Find the absolute maximum and minimum values of f (x, y) = x2 + y 2 + x2 y + 4

on D = {(x, y)| |x| ≤ 1, |y| ≤ 1}.

∇f = h2x + 2xy, 2y + x2 i
∇f = 0 ⇔ 2x + 2xy = 0, 2y + x2 = 0 ⇔ x(1 + y) = 0, 2y + x2 = 0
From x(1 + y) = 0, x = 0 or y = −1.
If x = 0, then 2y = 0 ⇒ y = 0 and (x, y) = (0, 0).
√ √ √
If y = −1, then −2 + x2 = 0 ⇒ x = ± 2. So (x, y) = ( 2, −1) or (− 2, −1).
√ √
Note that ( 2, −1), (− 2, −1) are on the outside of D. So on the interior of D,
there is a unique critical point (0, 0).

fxx = 2 + 2y, fxy = 2x, fyy = 2

At (0, 0),
2 0
D= =4>0
0 2
and fxx (0, 0) = 2 > 0. So f (0, 0) = 4 is a local minimum.
The boundary of D consists of four intervals, 1) x = 1, −1 ≤ y ≤ 1, 2) x =
−1, −1 ≤ y ≤ 1, 3) y = 1, −1 ≤ x ≤ 1, and 4) y = −1, −1 ≤ x ≤ 1.
On x = 1, f (x, y) = f (y) = y 2 + y + 5.
1
f 0 (y) = 2y + 1 = 0 ⇒ y = −
2

3
MATH 2004 Homework Solution Han-Bom Moon

1 19
f (− ) = , f (−1) = 5, f (1) = 7
2 4
19
So on −1 ≤ y ≤ 1, the maximum is 7, the minimum is .
4
2
On x = −1, f (x, y) = f (y) = y + y + 5. Therefore the maximum and minimum
values are same with the case of x = 1.
On y = 1, f (x, y) = f (x) = 2x2 + 5.

f 0 (x) = 4x = 0 ⇒ x = 0

f (0) = 5, f (−1) = f (1) = 7

So the maximum is 7, the minimum is 5.
Finally, on y = −1, f (x, y) = 5.
Therefore the absolute maximum is f (1, 1) = f (−1, 1) = 7, the absolute minimum
is f (0, 0) = 4.

14.7.41 Find the points on the cone z 2 = x2 + y 2 that are closest to the point (4, 2, 0).
(Note: You can solve this problem by using Lagrange multiplier. But I will show
a solution using the method in Section 14.7.)
In this problem, obviously the closest point exists. So the minimum among values
at critical points is the absolute minimum.
Distance from (x, y, z) to (4, 2, 0)
p p
= d = (x − 4)2 + (y − 2)2 + z 2 = x2 − 8x + y 2 − 4y + 20 + z 2
p p
= x2 − 8x + y 2 − 4y + 20 + x2 + y 2 = 2x2 − 8x + 2y 2 − 4y + 20
Let f (x, y) = d2 = 2x2 − 8x + 2y 2 − 4y + 20.

∇f = hfx , fy i = h4x − 8, 4y − 4i

∇f = 0 ⇒ 4x − 8 = 0, 4y − 4 = 0 ⇒ x = 2, y = 1
√ √
So (x, y, z) = (2, 1, ± 22 + 12 ) = (2, 1, ± 5) are the critical points. At these
two points, the distances to (4, 2, 0) are same. So both of them are local (and
the absolute) minimum.

14.7.55 Suppose that a scientist has reason to believe that two quantities x and y are
related linearly, that is y = mx + b, at least approximately, for some values of
m and b. The scientist performs an experiment and collects data in the form of
points (x1 , y1 ), (x2 , y2 ), · · · , (xn , yn ), and then plots these points. The points don’t
lie exactly on a straight line, so the scientist wants to find constants m and b so
that the line y = mx + b “fits” the points as well as possible.
Let di = yi − (mxi + b) be the vertical deviation of the point (xi , yi ) from the line.
Xn
The method of least squares determines m and b so as to minimize d2i , the
i=1

4
MATH 2004 Homework Solution Han-Bom Moon

sum of the squares of these deviations. Show that, according to this method, the
line of best fit is obtained when
n
X n
X
m xi + bn = yi ,
i=1 i=1

n
X n
X n
X
m x2i + b xi = xi yi .
i=1 i=1 i=1

Thus the line is found by solving these two equations in the two unknowns m
and b.

n
X n
X
f (m, b) = d2i = (yi − mxi − b)2
i=1 i=1

At (m, b) minimize f (m, b), ∇f = 0.

n
X n
X
∇f = hfm , fb i = h 2(yi − mxi − b)(−xi ), 2(yi − mxi − b)(−1)i
i=1 i=1

n
X n
X
= h−2 (yi − mxi − b)xi , −2 (yi − mxi − b)i
i=1 i=1
n
X n
X
∇f = 0 ⇒ (yi − mxi − b)xi = 0, (yi − mxi − b) = 0
i=1 i=1
n
X n
X n
X n
X
(yi − mxi − b)xi = 0 ⇒ yi xi − m x2i −b xi = 0
i=1 i=1 i=1 i=1
n
X n
X n
X
⇒m x2i +b xi = xi yi
i=1 i=1 i=1
n
X n
X n
X n
X
(yi − mxi − b) = 0 ⇒ yi − m xi − b=0
i=1 i=1 i=1 i=1
n
X n
X n
X n
X n
X
⇒m xi + b= yi ⇒ m xi + nb = yi
i=1 i=1 i=1 i=1 i=1

14.8.3 Use Lagrange multipliers to find the maximum and minimum values of f (x, y) =
x2 + y 2 subject to the given constraint xy = 1.

g(x, y) = xy
∇f = hfx , fy i = h2x, 2yi
∇g = hgx , gy i = hy, xi
∇f = λ∇g, g = 1 ⇔ 2x = λy, 2y = λx, xy = 1
λy λy
x= ⇒ 2y = λ · = λ2 y ⇒ 4y = λ2 y ⇒ (4 − λ2 )y = 0
2 2

5
MATH 2004 Homework Solution Han-Bom Moon

⇒ y = 0 or λ = ±2
y = 0 ⇒ xy = 0 so it is impossible.

λ = 2 ⇒ x = y ⇒ x = y = 1 or x = y = −1

⇒ (x, y) = (1, 1), (−1, −1)

λ = −2 ⇒ x = −y ⇒ xy = −x2 < 0
So it is impossible, too.
Therefore there are at most two local extrema at (1, 1) and (−1, −1). And f (1, 1) =
f (−1, −1) = 2.
Note that if x is very large positive number, f (x, y) = x2 + y 2 becomes very large
without any bound. So it has no absolute maximum. But its absolute minimum
is f (1, 1) = f (−1, −1) = 2. (Note that xy = 1 is not bounded so there may not be
an absolute maximum.)

14.8.5 Use Lagrange multipliers to find the maximum and minimum values of f (x, y) =
1
y 2 − x2 subject to the given constraint x2 + y 2 = 1.
4
1
g(x, y) = x2 + y 2
4
∇f = hfx , fy i = h−2x, 2yi
x
∇g = hgx , gy i = h , 2yi
2
x 1
∇f = λ∇g, g = 1 ⇒ −2x = λ , 2y = λ2y, x2 + y 2 = 1
2 4
⇒ y = λy ⇒ λ = 1 or y = 0
x 1
λ = 1 ⇒ −2x = ⇒ x = 0 ⇒ x2 + y 2 = 1 ⇒ y 2 = 1 ⇒ y = ±1
2 4
⇒ (x, y) = (0, 1) or (0, −1)
1 1
y = 0 ⇒ x2 + y 2 = 1 ⇒ x2 = 1 ⇒ x = ±2
4 4
⇒ (x, y) = (2, 0) or (−2, 0)
f (0, 1) = 1, f (0, −1) = 1, f (2, 0) = −4, f (−2, 0) = −4
So the absolute maximum is f (0, 1) = f (0, −1) = 1, the absolute minimum is
f (2, 0) = f (−2, 0) = −4.

14.8.10 Use Lagrange multipliers to find the maximum and minimum values of f (x, y, z) =
x2 y 2 z 2 subject to the given constraint x2 + y 2 + z 2 = 1.
First of all, because f (x, y, z) = x2 y 2 z 2 ≥ 0, it is easy to see that the absolute
minimum is 0 at (x, y, z) with at least one of x, y, and z is zero. So let’s find a local
maximum, minimum with x, y, z 6= 0.

g(x, y, z) = x2 + y 2 + z 2

6
MATH 2004 Homework Solution Han-Bom Moon

∇f = hfx , fy , fz i = h2xy 2 z 2 , 2yx2 z 2 , 2zx2 y 2 i

∇g = hgx , gy , gz i = h2x, 2y, 2zi
∇f = λ∇g, g = 1 ⇒ 2xy 2 z 2 = λ2x, 2yx2 z 2 = λ2y, 2zx2 y 2 = λ2z, x2 +y 2 +z 2 = 1
2xy 2 z 2 = λ2x ⇒ x = 0 or y 2 z 2 = λ
Now suppose that x, y, z 6= 0. Then by a similar compuation,

λ = y 2 z 2 = x2 z 2 = x2 y 2 ⇒ x2 = y 2 = z 2
1
⇒ x2 + y 2 + z 2 = 1 ⇒ x2 = y 2 = z 2 =
3
1
f (x, y, z) = x2 y 2 z 2 =
27
1
Therefore the absolute maximum is , the absolute minimum is 0.
27
14.8.20 Find the extreme values of f (x, y) = 2x2 + 3y 2 − 4x − 5 on the region described
by the inequality x2 + y 2 ≤ 16.

∇f = hfx , fy i = h4x − 4, 6yi

∇f = 0 ⇒ 4x − 4 = 0, 6y = 0 ⇒ (x, y) = (1, 0)
fxx = 4, fxy = 0, fyy = 6
At (1, 0),
4 0
D= = 24 > 0, fxx (1, 0) = 4 > 0.
0 6
So f (1, 0) = −7 is a local minimum.
Now the boundary is g(x, y) = x2 + y 2 = 16.

∇g = hgx , gy i = h2x, 2yi

∇f = λ∇g, g = 16 ⇒ 4x − 4 = λ2x, 6y = λ2y, x2 + y 2 = 16

6y = λ2y ⇒ λ = 3 or y = 0
If λ = 3, √
4x − 4 = 6x ⇒ x = −2 ⇒ x2 + y 2 = 16 ⇒ y = ± 12
√ √
(x, y) = (−2, 12) or (−2, − 12)
If y = 0,
x2 + y 2 = 16 ⇒ x = ±4
(x, y) = (4, 0) or (−4, 0)
√
f (−2, ± 12) = 47, f (4, 0) = 11, f (−4, 0) = 43
√
Therefore the absolute maximum is f (−2, ± 12) = 47, the absolute minimum is
f (1, 0) = −7.

7
MATH 2004 Homework Solution Han-Bom Moon

14.8.27 Use Lagrange multipliers to prove that the rectangle with maximum area that
has a given perimeter p is a square.
Let x be the width and y be the height of the rectangle. Then g(x, y) = 2x + 2y = p
and the area is A = xy.

∇A = hAx , Ay i = hy, xi
∇g = hgx , gy i = h2, 2i
At the maximum,

∇A = λ∇g, g = p ⇒ y = λ2, x = λ2, 2x + 2y = p.

⇒ x = 2λ = y
So the rectangle should be a square.

14.8.32 Find the points on the surface y 2 = 9 + xz that are closest to the origin.
Note that in this problem obviously there exists an absolute minimum.
Let f (x, y, z) be the distance square from the origin to (x, y, z). Then

f (x, y, z) = d2 = x2 + y 2 + z 2 .

g(x, y, z) = y 2 − 9 − xz = 0
We are finding the point (x, y, z) giving the minimum of f (x, y, z) on g(x, y, z) =
0.
∇f = hfx , fy , fz i = h2x, 2y, 2zi
∇g = hgx , gy , gz i = h−z, 2y, −xi
∇f = λ∇g, g = 0 ⇒ 2x = −λz, 2y = λ2y, 2z = −λx, y 2 − 9 − xz = 0
2y = λ2y ⇒ y = 0 or λ = 1
If y = 0,

λ λ2
2x = −λz ⇒ x = − z ⇒ 2z = −λx = z ⇒ z = 0 or λ = ±2
2 2
z = 0 is impossible because y 2 − 9 − xz = 0, y = 0 ⇒ xz = −9.

λ = 2 ⇒ x = −z ⇒ −9 + x2 = 0 ⇒ x = ±3

(x, y, z) = (3, 0, −3) or (−3, 0, 3)

λ = −2 ⇒ x = z ⇒ −9 − x2 = 0
This is impossible because the left hand side is negative.
If λ = 1,

2x = −z, 2z = −x ⇒ x = z = 0 ⇒ y 2 − 9 = 0 ⇒ y = ±3

8
MATH 2004 Homework Solution Han-Bom Moon

(x, y, z) = (0, 3, 0) or (0, −3, 0)

f (3, 0, −3) = f (−3, 0, 3) = 18, f (0, 3, 0) = f (0, −3, 0) = 9

Therefore the absolute minimum is 9.

14.8.36 Find the dimensions of the box with volume 1000 cm3 that has minimal surface
area.
Let x, y, z be the length, width, and height of the box respectively. Then V =
xyz = 1000. The surface area is A = 2xy + 2yz + 2zx.
∇A = hAx , Ay , Az i = h2y + 2z, 2x + 2z, 2x + 2yi
∇V = hVx , Vy , Vz i = hyz, xz, xyi
From xyz = 1000, we know that none of x, y, z is zero.
∇A = λ∇V, V = 1000 ⇒ 2y+2z = λyz, 2x+2z = λxz, 2x+2y = λxy, xyz = 1000
2 2 2 2 2 2
⇒ + = λ, + = λ, + = λ
y z z x y x
2 2 2 2 2 2 2 2 2
⇒ + = + = + ⇒ = = ⇒x=y=z
y z z x y x x y z
√
3
xyz = 1000 ⇒ x = y = z = 1000
√ √ √ 2
The minimum surface area is A( 1000, 1000, 1000) = 6(1000) 3 = 600 (cm2 ).
3 3 3

14.8.41 If the length of the diagonal of a rectangular box must be L, what is the largest
possible volume?
Let
p x, y, z be the length, width, and height of the box, respectively. Then L =
x2 + y 2 + z 2 . Let g(x, y, z) = x2 + y 2 + z 2 = L2 . The volume is V = xyz.

∇V = hVx , Vy , Vz i = hyz, xz, xyi

∇g = hgx , gy , gz i = h2x, 2y, 2zi
∇V = λ∇g, g = L2 ⇒ yz = λ2x, xz = λ2y, xy = λ2z, x2 + y 2 + z 2 = L2
⇒ xyz = 2λx2 , xyz = 2λy 2 , xyz = 2λz 2
⇒ 2λx2 = 2λy 2 = 2λz 2 ⇒ λ = 0 or x2 = y 2 = z 2
If λ = 0,
yz = 0 ⇒ y = 0 or z = 0 ⇒ V = xyz = 0
So obviously it does not give the absolute maximum.
If x2 = y 2 = z 2 ,
L2
3x2 = L2 ⇒ x2 = y 2 = z 2 = .
3
In this case,
s 3 r
p L2 L6 L3
V = xyz = x2 y 2 z 2 = = =√ .
3 27 27