Math 230.

01, Fall 2012: HW 1 Solutions

Problem 1 (p.9 #2). Suppose a word is picked at random from this sentence. Find:
a) the chance the word has at least 4 letters;
SOLUTION: All words are equally likely to be chosen. The sentence has 10 words; 7 are
at least 4 letters long. We have P(at least 4 letters) =
7
10
.
b) the chance that the word contains at least 2 vowels (a,e,i,o,u);
SOLUTION: Assuming that the vowels need not be distinct (i.e. the three instances of e
in sentence count as three vowels in the word), then exactly 4 words contain at least two
vowels, so P(at least 2 vowels) =
4
10
.
c) the chance that the word contains at least 4 letters and at least 2 vowels.
SOLUTION: Exactly four words contain at least 4 letters and at least two vowels, so P(at
least 4 letters and at least 2 vowels) =
4
10
.
d) What is the distribution of the length of the word picked?
SOLUTION: Let i be the length of the word. Then the values of i which occur with nonzero
probability are i = 1, 2, 4, 6, 7, and 8. We have P(1) =
1
10
, P(2) =
2
10
, P(4) =
3
10
, P(6) =
2
10
,
P(7) =
1
10
, P(8) =
1
10
.
e) What is the distribution of the number of vowels in the word?
SOLUTION: Let j be the number of vowels in the word. Then the values of j which oc-
cur with nonzero probability are j = 1, 2, 3. We have P(1) =
6
10
, P(2) =
2
10
, P(3) =
2
10
.
Problem 2 (p.9 #7). Suppose two 4-sided dice are rolled. Find the probabilities of the following
events:
a) the maximum of the two numbers rolled is less than or equal to 2;
SOLUTION: The sample space is the set of ordered pairs (i, j), where i and j are integers
between 1 and 4, and i represents the number on the upper face of the rst die and j the
number on the upper face of the second die. All outcomes are equally likely. There are 16
outcomes. The outcomes (1, 1), (1, 2), (2, 1), and (2, 2) are the outcomes whose maximum is
less than or equal to 2. Hence P(maximum is less than or equal to 2) =
1
4
.
b) the maximum of the two numbers rolled is less than or equal to 3;
SOLUTION: In order for the maximum of the two numbers to be less than or equal to 3,
1
both numbers must be less than or equal to 3. There are 3 3 = 9 ways for this to happen,
so P(maximum is less than or equal to 3) =
9
16
.
c) the maximum of the two numbers rolled is equal to 3;
SOLUTION: In order for the maximum to be equal to 3, at least one number must be 3,
and the another number can be anything less than or equal to 3. There are two choices for
which die results in a 3; once this is xed, there are 2 choices for the outcome of second die to
be strictly less than 3, yielding 2(3 1) outcomes in which the one number is 3 and the other
is strictly less than 3. Finally there is the outcome in which both numbers are 3. So there
are 231 outcomes in which the maximum is exactly 3, so the probability of this event is
5
16
.
d) Repeat part c for the maximum equal to 1, 2, and 4.
SOLUTION: The solution to this part is identical to the solutions above; P(max = 1) =
1
16
,
P(max = 2) =
3
16
, P(max = 4) =
7
16
.
e) If M is the maximum of the two numbers, then
P(M = 1) + P(M = 2) + P(M = 3) + P(M = 4) = 1,
check that your answers for c) and d) satisfy this relationship.
SOLUTION: Indeed, if the maximum is equal to i, then the upper face of at least one die
must be i (there are two possibilities for which die will have i on its upper face) and the other
die must result in a number strictly less than i or exactly equal to i. Now, there are (i-1)
numbers strictly less than i, and 1 number exactly equal to i, so we get 2(i 1) +i = 2(i) 1
possibilities for a maximum exactly equal to i, where i ranges from 1 to 4. Hence P(i) =
2i1
16
,
and evaluating this at each of i = 1, 2, 3 and 4 and then summing gives the desired result.
Problem 3. Now suppose that the die has n sides:
a) the maximum of the two numbers rolled is less than or equal to 2;
SOLUTION: Again, there are n
2
possible outcomes, all equally likely, and 4 outcomes in
which the maximum of the two numbers is less than 2, so the desired probability is
4
n
2
.
b) the maximum of the two numbers rolled is less than or equal to i {1, . . . n};
SOLUTION: As in the previous problem, for the maximum to be less than or equal to i,
both die rolls must result in numbers less than or equal to i. There are i choices for the
outcome of the rst die and i choices for the second, so i
2
ways in total for the maximum of
the two numbers to be less than i. The desired probability is
i
2
n
2
.
c) the maximum of the two numbers rolled is equal to i {1, . . . , n};
2
SOLUTION: As in the previous problem, one of the numbers must be equal to i and the
other must be less than or equal to i. If the rst one is equal to i, the second number can take
on i values; if the second is equal to i, the rst can take on i values, and the ordered pair (i, i)
should not be counted twice. Hence there are 2i 1 ways for the maximum to be equal to i.
The desired probability is therefore
2i1
n
2
.
Problem 4. Consider that the data
{.1, .4, 1.2, 1.22, 1.75, 2.1, 2.5, 2.6, 2.9, 2.99, 3.1415, 3.5, 3.3, 3.7}
and the cut points {0, 1, 2, 3, 4}. Draw the empirical distribution for this data. Label the height for
each rectangle as in Section 1.3. What is the total area of all of the rectangles?
SOLUTION. We note that the dierence between each consecutive pair of cut points (b
j
, b
j+1
)
is 1, and there are 14 data points in total. For each pair (b
j
, b
j+1
) of cut points, where the height of
the rectangle whose base is the interval between (b
j
, b
j+1
) is given by
H
j
=
# of data points x
i
for which b
j
< x
i
< b
j+1
[total number of data points][b
j+1
b
j
]
.
So, for instance, the height above the interval (2, 3) is
5
14
. The total area of all the rectangles sums
to 1, because the height of each rectangle is the fraction of data points that lie in the given interval
and the width of each rectangle is 1.
Problem 5 (p.31 #10). Events A, B, and C are dened on an outcome space. Find expressions for
the following probabilities in terms of P(A), P(B), P(C), P(AB), P(AC), P(BC), and P(ABC).
a) The probability that exactly two of the A, B, C occur.
SOLUTION: We want to nd the probability of the events E
1
= A B C
c
= ABC
c
,
E
2
= A C B
c
, and E
3
= B C A
c
. Each of these events E
i
represents exactly two of
A, B, or C occurring. Note that these events have no intersection. Therefore,
P ({A B C
c
} {A C B
c
} {B C A
c
}) = P(E
1
) + P(E
2
) + P(E
3
).
Now, P(E
1
) = P(AB) P(ABC), P(E
2
) = P(AC) P(ABC), and P(E
3
) = P(BC)
P(ABC). Hence we get
P(exactly two of the events occur) = P(AB) + P(AC) + P(BC) 3P(ABC).
b) The probability that exactly one of these events occurs.
SOLUTION: Here we want to nd the probabilities of the events F
1
= AB
c
C
c
, F
2
= BA
c
C
c
,
and F
3
= CA
c
B
c
. Again, the events F
i
are mutually exclusive and their union is event that
exactly one of the events occurs. We nd
P(F
1
) = P(A) P(AB) P(AC) + P(ABC);
P(F
2
) = P(B) P(AB) P(BC) + P(ABC);
P(F
3
) = P(C) P(AC) P(BC) + P(ABC)
which yields P(A) + P(B) + (C) 2P(AB) 2P(BC) 2(AC) + 3P(ABC) as the answer.
3
c) The probability that none of these events occur.
SOLUTION: The probability that no event occurs is the complement of the probability that
at least one event occurs. The probability that at least one event occurs is P(A B C). By
the inclusion-exclusion formula, we have
P(A B C) = P(A) + P(B) + P(C) P(AB) P(AC) P(BC) + P(ABC)
so the probability that none of the events occur is 1P(ABC), which by inclusion-exclusion
is
1 [P(A) + P(B) + P(C) P(AB) P(AC) P(BC) + P(ABC)] .
Problem 6. How many children should a family plan to have so that the probability of having at
least one child of each sex is at least 0.95? (Assume that both sexes are equally likely.)
SOLUTION: Let n be the number of children that the family will have, and let A
n
be the event that
there is at least one boy and one girl among the n children. It is easier to compute the probability
of the event, A
c
n
, that all n children are of the same sex. We want to nd n so that:
P(A
n
) = 1 P(A
c
n
) 0.95
P(A
c
n
) 0.05
Observe that the events M
n
= n boys, and F
n
= n girls are a partition of A
c
n
, and P(M
n
) =
P(F
n
) = 1/2
n
. So we solve:
P(A
c
n
) =
2
2
n
0.05
40 2
n
n log
2
40
The smallest value which satises this inequality is n = 6.
Problem 7. Suppose we roll two fair six-sided dice. What is the distribution of the dierence (in
absolute value) between the two numbers on the top faces?
SOLUTION: The smallest possible dierence is 0 and the largest is 6 1 = 5. When the dif-
ference is 0, the numbers must be equal there are 6 ways for this to occur. When the dierence is
1, there are 5 possibilities for the smaller number (6 cannot be the smaller of the two numbers, and
the larger number is determined by the smaller by adding 1), and for each possibility, there are two
outcomes (e.g. (1, 2) and (2, 1)), giving 10 ways for the dierence to be 1.
By similar reasoning, there are 4 2 = 8 outcomes with a dierence of 2, 3 2 = 6 outcomes
with a dierence of 3, and so on. The distribution is then
dierence, i 0 1 2 3 4 5
P(i)
6
36
=
1
6
5
18
2
9
1
6
1
9
1
18
The next problem is a classic puzzle.
4
Problem 8. Three people are going to play a cooperative game. They are allowed to strategize
before the game begins, but once it starts they cannot communicate with one another. The game
goes as follows. A fair coin is tossed for each player to determine whether that player will receive
a red hat or a blue hat, but the color of the hat (and result of the coin toss) is not revealed to the
player. Then the three players are allowed to see one another, so each player sees the other two
players hats, but not her own. Simultaneously, each player must either guess at the color of her
own hat or pass. They win if nobody guesses incorrectly and at least one person guesses correctly
(so they cant all pass).
The players would like to maximize their probability of winning, so the question is what should
their strategy be? A naive strategy is for them to agree in advance that two people will pass and one
person (designated in advance) will guess either red or blue. This strategy gives them a 50% chance
of winning, but it is not optimal. Devise a strategy that gives the players a greater probability of
winning.
SOLUTION: First, write out the outcome space for the assignment of hats.
= {RRR, RRB, RBR, BRR, RBB, BRB, BBR, BBB}
The goal is not to win for every outcome, but to win for more than half of the outcomes. Notice that
in all but two of the outcomes, two people have the same color hat and one person has a dierent
colored hat. In each of these scenarios, then, one person sees two like-colored hats, while the other
two people see one hat of each color. For example, if the outcome is RRB, then the person with the
blue hat sees two red hats on the other players heads, and each person with a red hat sees a red
and a blue hat. Therefore, the following strategy will guarantee a win whenever two people have the
same colored hat and one person has the other color: If a player sees two hats of the same color, then
guess the opposite color; If a player sees two dierent colored hats, then pass. This strategy always
fails when all three hats are the same color, since all three players will guess incorrectly. Therefore,
the probability of winning with this strategy is 6/8 = 75%.
Now for a little review of calculus. These are things you should have learned in a class before
this one. If you have problems with them, get help now to refresh your memory.
Problem 9. Do the following integrals:
_
1
0
x
3
dx
_
0

xexp(x)dx
_

exp(x
2
/2)dx
SOLUTION. For the rst integral, by the fundamental theorem of calculus and the power rule,
_
1
0
x
3
dx =
x
4
4

1
0
=
1
4
For the second integral, note that the integrand is the product of two functions: x and exp(x), one
of which gets simpler with dierentiation and the other of which is easy to integrate. Hence this
type of integral can be addressed through integration-by-parts.
Also, as x , xexp(x) 0 (write the expression as a quotient and use Lh opitals rule). We
derive
_
0

xexp(x)dx = [xexp(x)]|
0

_
0

exp(x)dx = 1
5
The last integral requires a bit of ingenuity. Let us rst observe that
_

exp(x
2
/2)dx =
_

exp(y
2
/2)dy;
that is, in a denite integral, the particular choice of letter to represent the variable of integration
is irrelevant. Hence we get
__

exp(x
2
/2)dx
_ __

exp(y
2
/2)dy
_
=
__

exp(x
2
/2)dx
_
2
Now, consider the double integral
_

exp
_
(x
2
+ y
2
)/2

dxdy
This is an iterated integralrst we integrate with respect to the variable x, and then with respect
the variable y. A priori, we cannot say that the iterated integral and the product of the two
separate integrals are equal. But since exp
_
(x
2
+ y
2
)/2

= exp(x
2
/2) exp(y
2
/2), the function
f(x, y) = exp
_
(x
2
+ y
2
)/2

can be decomposed into a product of functions f(x, y) = g(x)h(y),

where g(x) = exp(x
2
/2), and h(y) = exp(y
2
/2). While integrating f(x, y) = g(x)h(y) rst in the
variable x, we can treat h(y) as a constant. Namely, we have
_

exp
_
(x
2
+ y
2
)/2

dxdy =
_

exp(y
2
/2)
__

exp(x
2
/2)dx
_
dy
=
__

exp(x
2
/2)dx
_ __

exp(y
2
/2)dy
_
=
__

exp(x
2
/2)dx
_
2
From the above chain of equalities, we see that if we can evaluate the double integral over the whole
plane, then we can determine the value of the integral we want. The double integral over the whole
plane
_

exp
_
(x
2
+ y
2
)/2

dxdy
can be evaluated by changing to polar coordinates: x = r cos , y = r sin, where 0 r < ,
0 2. By the change-of-variables formula for multiple integrals, in a change of variables from
(x, y) to (r, ), the area dierential dxdy is transformed to the dierential drd multiplied by the
absolute value of the determinant of the Jacobian matrix of the change-of-coordinate map

(x,y)
(r,)

.
You can compute this matrix (in fact you should!), and you will nd its determinant to be r, which
is always nonnegative. Hence the area dierential dxdy is transformed under this change of variable
to rdrd. The double integral in the new (r, ) coordinates is
_

exp
_
(x
2
+ y
2
)/2

dxdy =
_
2
0
_

0
exp
_
r
2
/2

rdrd
Since the function being integrated does not depend on , this becomes
2
_

0
exp
_
r
2
/2

rdr
6
which can be evaluated by simple substitution. The nal answer for the double integral is 2, so
the nal answer for the original integral is

2.
Problem 10. Perform the following dierentiations:
d
dx
(x
4
)
d
dx
(x
2
exp(x))
d
dx
(ln(x
2
))
SOLUTION. These are straightforward and we omit the details: the rst dierentiation is an appli-
cation of the power rule; the second an application of the product rule; the third an application of
the chain rule. We get
d
dx
(x
4
) = 4x
3
d
dx
(x
2
exp(x)) = (2x x
2
) exp(x)
d
dx
(ln(x
2
)) =
2
x
Problem 11. Evaluate the following innite sums:

k=1
_
1
4
_
k

k=1
3
k
k!
SOLUTION. The rst series is the sum of a geometric series whose rst term is
1
4
and whose ratio
is
1
4
. Hence its sum is
1/4
11/4
=
1/4
3/4
= 1/3. The second series is almost the series expansion for the
exponential function exp(x), except the rst term in that series,
3
0
0!
= 1, is not included. Hence the
numerical value of the innite sum is exp(3) 1.
Problem 12. Draw a picture of the region of the xy-plane were both x and y are between 0 and 1
and y x
2
. Find the area of this region.
SOLUTION. The area is represented by the integral
_
1
0
[1 x
2
]dx = [x x
3
/3]

1
0
= 2/3
Problem 13. Evaluate the integral
_
4
0
_
2
y
2
e
x
2
dxdy
Hint: Draw the region over which you are integrating and change the order of integration.
SOLUTION. The region is {y/2 x 2, 0 y 4}. In order to change the order of integration, we
need to express the limits rst in terms of y (so the y limits depend on x), and then in terms of x,
so the x limits are then two numbers. We see that if y/2 x 2 and 0 y 4, then 0 y 2x,
and 0 x 2. (Draw these two regions separately and check that they are the same region!) By
interchanging the order of integration, we get
_
4
0
_
2
y
2
e
x
2
dxdy =
_
2
0
_
2x
0
e
x
2
dydx
=
_
2
0
2xe
x
2
dx
= exp(x
2
)

2
0
= exp(4) 1
7