Rosen, Discrete Mathematics and Its Applications, 6th edition
Extra Examples
Section 7.1—Recurrence Relations
p.451, icon at Example 3
#1. Find a recurrence relation (and initial condition) for each of the following:
(a) the number of strings of length n of letters of the alphabet.
(b) the number of strings of length n of letters of the alphabet, if no adjacent letters can be the same.
(c) the number of strings of length n of letters of the alphabet with no repeated letters.
Solution:
(a) Let an equal the number of strings of length n of letters of the alphabet. We can obtain any such string
by taking a string s of length n − 1 and appending a letter to the end of s. This can be done in 26 ways.
Therefore, an = 26an−1. The initial condition is a1 = 26.
(b) Let bn equal the number of strings of length n of letters of the alphabet with no adjacent letters identical.
Each such string can be obtained from a string s of length n − 1 by taking s and appending to it a letter
that is different from the last letter of s. Because there are 25 letters that can be appended, there are 25
ways to extend s to a string of length n. Therefore, bn = 25bn−1. The initial condition is b1 = 26.
(c) Let cn equal the number of strings of length n of letters of the alphabet with no repeated letters. Each
such string can be obtained from a string s of length n − 1 by taking s and appending to it a letter that is
different from each of the letters of s. Because there are n−1 letters used in s, there are 26−(n−1) = 27−n
letters available to be appended to s. Therefore, cn = (27 − n)cn−1. The initial condition is c1 = 26. Note
that c27 = (27 − 27)c26 = 0 because there are only 26 letters in the alphabet. Likewise, the recurrence
relation yields c28 = c29 = . . . = 0.
p.451, icon at Example 3
#3. Suppose bn = 2bn−1 + n − 2n and b0 = 5.
(a) Find bn−1 in terms of bn−2.
(b) Find bn in terms of bn−2.
(c) Find bn in terms of bn−3.
(d) Use parts (b) and (c) to conjecture a formula for bn.
Solution:
(a) The recurrence relation for bn doubles the previous term (which is 2bn−1), adds the subscript number of
bn (which is n), and subtracts 2 raised to the power of the subscript of bn (which is 2n).
The term bn−1 is obtained in the same way: double the previous term (which is 2bn−2), add the subscript
number of bn−1 (which is n−1), and subtract 2 raised to the power of the subscript of bn−1 (which is 2n−1).
Therefore bn−1 = 2bn−2 + (n − 1) − 2n−1.
(b) To obtain bn in terms of bn−2, we first use the recurrence relation to obtain bn in terms of bn−1 and then
use part (a) to obtain bn−1 in terms of bn−2:
bn = 2bn−1 + n − 2n
= 2[2bn−2 + (n − 1) − 2n−1] + n − 2n
= 22bn−2 + (n − 1) + n − 2n−1 − 2n.
(c) To obtain bn in terms of bn−3, we can use the recurrence equation for bn−2 and part (b). The recurrence
relation for bn−2 is bn−2 = 2bn−3 +(n−2)−2n−2. Substituting for bn−2 into the result in part (b), we have
bn = 22bn−2 + (n − 1) + n − 2n−1 − 2n
= 22[2bn−3 + (n − 2) − 2n−2] + (n − 1) + n − 2n−1 − 2n
= 23bn−3 + (n − 2) + (n − 1) + n − 2n−2 − 2n−1 − 2n.
(d) Mentally continuing the pattern in part (c), it seems reasonable to guess that the first term becomes
2nb0, the sum (n−2)+(n−1)+n becomes 1+2+3+· · ·+(n−2)+(n−1)+n, and the sum of the powers
of 2 being subtracted becomes −21 − 22 − 23 −· · ·−2n−1 − 2n−1 − 2n. Thus, it is reasonable to conjecture
that
bn = 2nb0 + [1 + 2 + 3 + · · · + (n − 2) + (n − 1) + n] − [21 + 22 + 23 + · · ·+ 2n−1 + 2n−1 + 2n]
= 5· 2n + n(n + 1)
2
− 2n+1,
2
where summation formulas were used at the last step to replace 1 + 2 + 3 + · · · + (n − 2) + (n − 1) + n and
21 + 22 + 23 + · · ·+ 2n−1 + 2n−1 + 2n.
Note: You can check that this is correct by substituting the formula for bn and bn−1 into the given recurrence
relation and showing that both sides of the recurrence relation are equal.
p.451, icon at Example 3
#4. Solve: an = 3an−1 + 1, a0 = 4, by substituting for an−1, then an−2, etc.
Solution:
Beginning with an = 3an−1 + 1 and substituting for an−1, then an−2, then an−3, etc., yields:
an = 3an−1 + 1
= 3(3an−2 + 1)+ 1
= 32an−2 + 3 · 1 + 1
= 32(3an−3 + 1) + 3 · 1 + 1
= 33an−3 + 32 · 1 + 3 · 1 + 1
...
= 3na0 + (3n−1 + 3n−2 + · · ·+ 32 + 3 + 1)
= 4· 3n +
3n − 1
2
=
9
2
· 3n − 1
2
=
3n+2
2
− 1
p.451, icon at Example 3
#5. Find a formula for the recurrence relation an = 2an−1 + 2n, a0 = 1, using a recursive method.
Solution:
an = 2an−1 + 2n
= 2(2an−2 + 2n−1) + 2n = (22an−2 + 2·2n−1) + 2n = (22an−2 + 2n) + 2n = 22an−2 + 2·2n
= 22(2an−3 + 2n−2) + 2·2n = (23an−3 + 22·2n−2) + 2·2n = (23an−3 + 2n) + 2·2n = 23an−2 + 3·2n.
At this stage it would be reasonable to guess that if we continue this process we will obtain:
an = 2na0 + n · 2n = 2n · 1 + n · 2n = (n + 1)2n.
Therefore, an = (n + 1)2n is a formula for the given sequence.
We can verify that this formula is correct by substituting an = (n+1)2n and an−1 = n2n−1 into the original
recurrence relation and checking that an equality results.:
2an−1 + 2n = 2(n2n−1) + 2n = n2n + 2n = (n + 1)2n = an.
p.451, icon at Example 3
#6. You begin with $1000. You invest it at 5% compounded annually, but at the end of each year you
withdraw $100 immediately after the interest is paid.
(a) Set up a recurrence relation and initial condition for the amount you have after n years.
(b) How much is left in the account after you have withdrawn $100 at the end of the third year?
(c) Find a formula for an.
(d) Use the formula to determine how long it takes before the last withdrawal reduces the balance in
the account to $0.
Solution:
(a) For n > 0, let an be the amount in the account at the end of year n; i.e., just after the interest has been
added to the account and the $100 has been withdrawn. Then an is equal to the amount from the previous
year (an−1) plus the interest earned (0.05an−1) minus the $100 withdrawal. That is,
an = an−1 + 0.05an−1 − 100 = 1.05an−1 − 100, if n > 0, a0 = 1000.
(b) Using the recurrence relation yields
a1 = 1050 − 100 = 950
a2 = 1.05a1 − 100 = 1.05(950)− 100 = 997.50 − 100 = 897.50
a3 = 1.05a2 − 100 = 1.05(897.50)− 100 = 942.38 − 100 = 842.38
Thus, the answer is $842.38.
(c) To develop a formula, note that
an = 1.05an−1 − 100,
an−1 = 1.05an−2 − 100,
an−2 = 1.05an−3 − 100,
...
Therefore
an = 1.05an−1 − 100
= 1.05(1.05an−2 − 100) − 100
= 1.052an−2 − (1.05·100) − 100
= 1.052(1.05an−3 − 100) − (1.05·100) − 100
= 1.053an−3 − (1.052·100) − (1.05·100) − 100
...
= 1.05na0 − (1.05n−1·100) − (1.05n−2·100)−· · ·−(1.052·100) − (1.05·100) − 100
= 1.05n·1000 − 100(1.05n−1 + 1.05n−2 + · · ·+ 1.052 + 1.05 + 1)
= 1.05n·1000 − 100 · 1.05n − 1
1.05 − 1
= 1.05n·1000 − 2000(1.05n − 1)
= 2000 − 1.05n·1000
and hence a formula for an is an = 2000 − 1.05n · 1000.
(d) Using various values of n in the formula for an yields a14 = 20.07 and a15 = −78.93. Hence, at the
end of the 15th year the balance will be 1.05 · 20.07 = 21.07 before a withdrawal is made; if this amount is
withdrawn, the balance will become $0. (Alternately, we could solve the equation 2000 − 1.05n · 1000 = 0
for n and obtain n as the solution.)
p.451, icon at Example 3
#7. Find a recurrence relation for the number of strings of letters of the ordinary alphabet that do not
have adjacent vowels.
Solution:
Let us call a string of letters of the alphabet “good” if it has no adjacent vowels. Let an be the number of
strings of length n of letters of the alphabet that do not have adjacent vowels.
The set of all good strings of length n is the union of the following two disjoint sets —
A: the set of good strings of length n that end with a consonant, and
B: the set of good strings of length n that end with a vowel.
Each string in A can be obtained from a good string of length n − 1 by adding any consonant at the end of
the string. Thus, |A| = 21 · an−1.
Each string in B ends with a vowel. In this case we know that that second letter from the end must be
a consonant (otherwise the string would have adjacent vowels). Thus, each string in B is a good string
of length n − 2 followed by a consonant (in the second last position) and a vowel at the end. Therefore,
|B| = 5· 21 · an−2.
Because A and B are disjoint, the set of all good strings of length n is their sum, |A| + |B|. That is,
an = 21an−1 + 105 an−2.
The two initial conditions are obtained by counting the number of good strings of lengths 1 and 2: a1 = 26
because any string of one of the 26 letters of the alphabet cannot have adjacent vowels; a2 = 262 −52 = 651
(we take all strings of length 2 and subtract those with two vowels).
p.451, icon at Example 3
#8. You have two distinct parallel lines L1 and L2. You keep adding additional lines, L3, L4, . . ., with
none parallel to L1 or L2 or to each other, and no three passing through the same point.
(a) Find a recurrence relation and initial condition(s) for rn, which is defined to be the number of regions
into which the plane is divided by the lines L1, L2, . . ., Ln.
(b) Find a formula for the number of regions into which the plane is divided by L1, L2, . . ., Ln.
Solution:
(a) The recurrence relation is rn = rn−1 + n (n > 2), with initial condition r2 = 3. To see this, note that
line Ln must cut each of the previous n − 1 lines in exactly one point. This in effect divides Ln into n
segments, each of which divides an existing region into two parts. Therefore, rn = rn−1 + n.
(b) Proceeding inductively:
r2 = 3,
r3 = r2 + 3 = 3 + 3,
r4 = r3 + 4 = 3 + 3 + 4,
r5 = r4 + 5 = 3 + 3 + 4 + 5,
r6 = r5 + 6 = 3 + 3 + 4 + 5 + 6.
This suggests that
rn = 3 + (3 + 4 + · · · + n)
= (1 + 2) + (3 + 4 + · · · + n)
= 1 + 2 + 3 + · · · + n
= n(n + 1)
2 .
To verify that this guess is correct, take rn = rn−1 + n and substitute n(n + 1)
2
for rn and
(n − 1)n
2
for
rn−1, obtaining n(n + 1)
2
=
(n − 1)n
2
+n, which is true because the right side simplifies to give the left side:
(n − 1)n
2
+ n =
(n − 1)n + 2n
2
= n2 + n
2
= n(n + 1)
2
.
p.451, icon at Example 3
#9. This is a variation on Fibonacci’s rabbit sequence. We begin with one pair of newborn rabbits. Once a
pair is two months old, the pair has two pairs of offspring, and continues to have two pairs of offspring each
month thereafter. Give a recurrence relation and initial condition(s) for the sequence fn, where fn is equal
to the number of pairs of rabbits alive at the end of the nth month (after the offspring are born). Assume
that the rabbits never die during the period being considered.
Solution:
The number of pairs at the end of n months (fn) is equal to the number of pairs alive one month earlier
(fn−1) plus the number of newborn pairs. But each pair of rabbits alive two months earlier gives birth to
two pairs of newborn rabbits. Therefore, there are 2fn−2 pairs of newborn rabbits. Hence
fn = fn−1 + 2fn−2, f(1) = 1, f(2) = 3.
p.451, icon at Example 3
#10. Here is another variation on Fibonacci’s rabbit sequence. We begin with one pair of newborn rabbits.
At the end of each month a new pair of newborn rabbits is added to the population. Once any pair is
two months old, the pair has one pair of offspring and continues to have one pair of offspring each month
thereafter. Give a recurrence relation and initial condition(s) for the sequence fn, where fn is equal to the
number of pairs of rabbits alive at the end of the nth month (after the rabbits have given birth and the
newborn pair has been introduced). Assume that the rabbits never die during the period being considered.
Solution:
The number of pairs at the end of n months (fn) is equal to 1 (the newborn pair added) plus the number
of pairs alive one month earlier (fn−1) plus the number of newborn pairs (which is equal to the number of
pairs alive two months earlier, fn−2. Thus
fn = 1+fn−1 + fn−2, f(1) = 2, f(2) = 4.
p.451, icon at Example 3
#11. Here is a third variation on Fibonacci’s rabbit sequence. We begin with one pair of newborn rabbits.
Once the pair is three months old, the pair has one pair of offspring, and continues to have one pair of offspring
every other month thereafter. Give a recurrence relation and initial condition(s) for the sequence fn, where fn
is equal to the number of pairs of rabbits alive at the end of the nth month (just after any offspring are
born). Assume that the rabbits never die during the period being considered.
Solution:
The number of pairs alive at the end of n months (fn) is equal to the number of pairs alive one month earlier
(fn−1) plus the number of pairs of newborn rabbits. To determine the number of newborn pairs, we need to
know the number of pairs born three months earlier, five months earlier, seven months earlier, etc. But, for
example, the number of pairs born three months before month n is equal to fn−3 −fn−4 and the number of
pairs born five months before month n is equal to fn−5 − fn−6.
Thus,
fn = fn−1 + number of pairs born in month n
= fn−1 + (fn−3 − fn−4) + (fn−5 − fn−6) + (fn−7 − fn−8) + · · ·
where the sum continues as long as the subscripts are positive. Similarly,
fn−1 = fn−2 + number of pairs born in month n − 1
= fn−2 + (fn−4 − fn−5) + (fn−6 − fn−7) + (fn−8 − fn−9) + · · · .
Substitute fn−1 from the second equation into the first equation. Almost all terms cancel and we obtain
fn = fn−2 + fn−3,
with f(1) = 1, f(2) = 1, f(3) = 2.
p.451, icon at Example 3
#12. (Problem A1 from the 1990 William Lowell Putnam Mathematics Competition)
Here are the first ten terms of an infinite sequence:
2, 3, 6, 14, 40, 152, 784, 5168, 40567, 363392.
(a) Find a formula for an infinite sequence a0, a1, a2, a3, . . . such that the first ten terms of the sequence are
the ones given here. (Hint: consider the sum of two rapidly increasing sequences.)
(b) Show that the sequence in (a) satisfies the recurrence relation
an = (n + 4)an−1 − 4nan−2 + (4n − 8)an−3.
Solution:
(a) The sequence increases rapidly, which suggests the possibility that the formula may involve an exponential
function cn. If we look for 2n in each term, we find the the first few terms are:
20 + 1, 21 + 1, 22 + 2, 23 + 6, 24 + 24, 25 + 120, 26 + 720, 27 + 5040 .
The second term in each sum is a factorial, yielding
20 + 0!, 21 + 1!, 22 + 2!, 23 + 3!, 24 + 4!, 25 + 5!, 26 + 6!, 27 + 7! .
Thus, an = 2n + n! is one such formula.
(b) We need to show that an = (n+4)an−1−4nan−2+(4n−8)an−3 is satisfied by the sequence an = 2n+n!.
That is,
2n + n! = (n + 4)[2n−1 + (n − 1)! ] − 4n[ 2n−2 + (n − 2)! ] + (4n − 8)[ 2n−3 + (n − 3)! ].
The right side can be simplified as follows:
(n + 4)[2n−1 + (n − 1)! ] − 4n[ 2n−2 + (n − 2)! ] + (4n − 8)[ 2n−3 + (n − 3)! ]
= n2n−1 + 4 · 2n−1 − 4n2n−2 + 4n2n−3 − 8 · 2n−3+
n(n − 1)! + 4(n − 1)! − 4n(n − 2)! + (4n − 8)(n − 3)!
= n2n−1 + 2n+1 − n2n + n2n−1 − 2n + n! + 4(n − 1)! − 4n(n − 2)! + (4n − 8)(n − 3)!
= n2n + 2n+1 − n2n − 2n + n! + 4(n − 1)! − 4n(n − 2)! + (4n − 8)(n − 3)!
= 2n+1 − 2n + n! + 4(n − 1)! − 4n(n − 2)! + (4n − 8)(n − 3)!
= 2n + n! + 4(n − 1)! − 4n(n − 2)! + (4n − 8)(n − 3)!
= 2n + n! + 4(n − 1)! − 4n(n − 2)! + 4(n − 2)!
= 2n + n! + 4(n − 1)! + (4 − 4n)(n − 2)!
= 2n + n! + 4(n − 1)! − 4(n − 1)(n − 2)!
= 2n + n! + 4(n − 1)! − 4(n − 1)!
= 2n + n!.
p.451, icon at Example 3
#13. Suppose a chess king is placed on the lower left square of an m×n chessboard (that is, a rectangular
board with m rows and n columns). Let M(m, n) be equal to the number of paths that a king can use
moving from the lower left corner to the upper right corner of an m×n board, with the restriction that each
move is either up, to the right, or diagonally up and to the right.
(a) Find a recurrence relation and initial condition(s) for M(m, n).
(b) Find the number of ways in which the king can move from the lower left square to the upper right
square on a 5 × 5 chessboard.
Solution:
(a) In order for the king to reach the upper right square, the king’s last move must be from one of the
following three squares: the square immediately below the corner square, the square immediately to the left
of the corner square, or the square diagonally down and to the left of the corner square. The number of
ways in which the king could have arrived at each of these three squares is M(m, n − 1), M(m − 1, n), and
M(m− 1, n − 1), respectively (assuming that m > 1 and n > 1). Therefore,
M(m, n) = M(m, n − 1) +M(m− 1, n) +M(m− 1, n − 1),
with initial conditions M(1, 1) = M(2, 1) = M(1, 2) = 1.
(b) First note that M(k, 1) = M(1, k) = 1 for all k > 1 and M(j, k) = M(k, j) for all j and k (by symmetry).
The following steps use the recurrence relation to find M(5, 5):
M(2, 2) = M(2, 1) +M(1, 2) +M(1,1) = 1 + 1 + 1 = 3
M(2, 3) = M(3, 2) = M(1, 3) +M(2, 2) +M(1,2) = 1 + 3 + 1 = 5
M(3, 3) = M(3, 2) +M(2, 3) +M(2,2) = 5 + 5 + 3 = 13
M(2, 4) = M(4, 2) = M(1, 4) +M(2, 3) +M(1,3) = 1 + 5 + 1 = 7
M(2, 5) = M(5, 2) = M(1, 5) +M(2, 4) +M(1,4) = 1 + 7 + 1 = 9
M(3, 4) = M(4, 3) = M(3, 3) +M(2, 4) +M(3,2) = 13 + 7 + 5 = 25
M(3, 5) = M(5, 3) = M(4, 3) +M(5, 2) +M(2,4) = 25 + 9 + 7 = 41
M(4, 4) = M(4, 3) +M(3, 4) +M(3, 3) = 25 + 25 + 13 = 63
M(4, 5) = M(5, 4) = M(4, 4) +M(3, 5) +M(3, 4) = 63 + 41 + 25 = 129
M(5, 5) = M(5, 4) +M(4, 5) +M(4, 4) = 129 + 129 + 63 = 321.
Therefore, M(5, 5) = 321.
This is shown in the following table, where, beginning from the lower left corner, each square has as its value
the sum of the numbers in the squares directly below, to the left, and diagonally below on the left.
1 9 41 129 321
1 7 25 63 129
1 5 13 25 41
1 3 5 7 9
1 1 1 1 1
(arrow means move, circle means scoop).
Q, the proposition Q 
Q 
X], [X 
Y], [X 
Y], [X 
Y], and [X 
Y]
, 
, 
, 
, 
,