A recurrence relation is an equation that specifies one (or more) known terms in a sequence and then a rule for calculating the next term in the sequence. For example:
\text{$t_{n+1}=t_{n}+d$, where $t_n=a$ and $n=0,1,2,...$}
Where a is the value of the first term, and d is some constant being added each time. The same recurrence relation can be expressed as follows:
\text{$t_n = t_{n-1}+d$, where $t_{n-1} = a$ and $n= 1,2,3,...$}
When considering a general sequence t_1, t_2, t_3, ..., t_{n-1}, t_{n}, t_{n+1}, a recurrence rule can be defined such that the nth term t_n or the n+1^{\text{th}} term t_{n+1} gives the next term in the sequence.
First order linear recurrence relations are recurrence equations that have the following form:
t_{n+1} = Rt_n + d, t_0 = a where a,R, and d are given constants.
It can sometimes be expressed as t_n in terms of t_{n-1}.
t_n = rt_{n-1} + d, t_1 = a where a,r, and d are given constants.
The relation above is known as a first order recurrence relation, because the rule only requires one single previous term in the sequence to generate the next term in the sequence. It is also a linear recurrence relation, because the previous referenced sequence terms are linear with r and d are given as constants.
Note that r could be a unit fraction such as \dfrac12 to represent the previous term being divided by constant 2. Also note that d could be a negative number such as -1, to show constant 1 being subtracted from the previous term. Therefore first order linear recurrence relations can use addition, subtraction, multiplication, and division operations to create sequences, not just addition and multiplication.
Consider the sequence defined by a_0=-8 and a_{n+1}=-\dfrac{1}{2}a_n for n\geq 0.
What is the first term of the sequence?
What is the second term of the sequence?
What is the third term of the sequence?
What is the fourth term of the sequence?
What is the fifth term of the sequence?
First order linear recurrence relations are recurrence equations that have the following form:
It can sometimes be expressed as t_n in terms of t_{n-1}.
An arithmetic sequence can be defined by recurrence relation t_{n+1}=t_{n}+d with t_0=a. The constant coefficient d represents the common difference between each term and the constant a represents the first term in the sequence.
Arithmetic sequences will be explored further in the  next lesson of this chapter .
Consider t_{n+1}=t_{n}-7, t_0=100. Describe the meaning of this rule in words.
Identify the recurrence relations that generate arithmetic sequences. Select all that apply.
Arithmetic sequence can be defined by the following recurrence relation.
It can sometimes be expressed as t_n in terms of t_{n-1}.
A geometric sequence can defined by recurrence relation t_{n+1}=Rt_{n} with t_0=a given as the first term and where each new term in the sequence is found by multiplying the previous term by a common ratio R.
Consider t_{n+1}=\dfrac{1}{2} t_{n}, t_0=128. Using the recurrence formula, the first few terms are 128, 64, 32, 16,... which is a geometric sequence, beginning with the first term a=128. Each new term is found by multiplying by a common ratio of R=\dfrac{1}{2}. In other words, the previous term in the sequence is divided by 2 to create the next term.
Geometric sequences will be explored further in a  later lesson within this chapter .
Identify the recurrence relations that generate geometric sequences. Select all that apply.
Geometric sequence can be defined by the following recurrence relation.
It can sometimes be expressed as t_n in terms of t_{n-1}.
Finally a recurrence equation in the form t_{n+1}=Rt_{n}+d with t_0=a, R \neq 1, and d \neq 0 forms a sequence that is neither geometric nor arithmetic, but a combination of both. t_{n+1}=3t_{n}+1, t_0=0 is an example of a sequence that is neither arithmetic nor geometric.
A sequence is neither arithmetic nor geometric if the recurrence relation is in the following form.
If a recurrence relation requires two or more previous terms in the sequence to generate the next term in the sequence, then it is considered a higher order recurrence relation.
For example, t_{n+2} = t_{n+1} + t_n where t_0 = 1 and t_1 = 1 is a second order recurrence relation, as the first two terms are required for future terms in the sequence to be generated. This particular second order linear recurrence relation is famously known as the Fibonacci sequence, which will be studied in a later lesson within this chapter.
A higher order recurrence relation requires two or more previous terms to find the next term.
A non-linear recurrence relation can also exist. A recurrence relation is considered non-linear when the recursive formula includes operations that are no longer limited to constants being added, subtracted or multiplied to previous terms.
Consider the recurrence relation t_{n+1} = t_n + n for when t_0 = 1. The previous term in the first relation has a number n being added, which is not a constant, the number being added gets larger and larger for later terms. t_1 = t_0 +0 = 1 + 0 = 1, then t_2 = 1 + 1 = 2, t_3 = 2 + 2 = 4 and so forth. Or consider t_{n+1} = (t_n)^2. The previous term in this relation is being multiplied by itself and therefore not a constant multiplier. These are both examples of non-linear recurrence relations.
In this chapter, only specific linear recurrence relations will be explored in further depth.
A recurrence relation is not linear if the recursive formula does not have a constant.
The graph of a recurrence relation can provide a visual understanding of the way the sequence behaves. For instance, it may show whether the terms are always increasing or always decreasing in size. It might also show that the terms change linearly, or as a quadratic curve, or as some other known form, and this might provide a hint as to a possible explicit form of the sequence.
Consider a stockbroker's claim that \$200 can be turned into a \$1000 in 10 months. The stockbroker claims that they can make 30\% a month on the investment, for the cost of \$35 a month.
Such a claim seems incredulous, but given the percentage earnings and the monthly fee stated, a recurrence relation can be formed as:
U_{n+1}=1.3U_n-35,U_0=200
Note that the coefficient 1.3 is equivalent to 130\%, which is a 30\% increase on the previous month's amount as shown by the formula. The subtraction of \$35 represents the monthly fee. The initial balance is \$200 so we call this U_0 therefore U_1 is the amount at the end of month 1.
The recurrence relation can be used to create a table of values, stating the size of the investment at the beginning of the month for the first 10 months as follows, using month numbers (with the first month referred to as month 0) and rounded whole dollars:
\text{M} | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
---|---|---|---|---|---|---|---|---|---|---|
\$ | 200 | 225 | 258 | 300 | 355 | 426 | 519 | 640 | 796 | 1000 |
The end of month 1 shows that the investment has grown to \$225, determined using the recurrence formula, so that U_1=1.3\times 200-35=225. Similarly, the end of month 2 is determined as U_2=1.3\times 225-35=257.50 which rounded to whole dollars becomes \$258.
Note the increase in the original investment, with the rate of change in value increasing as well. Although not strictly geometric (the \$35 deduction affects the ratio between successive terms), the trend is certainly apparent - the higher the amount, the faster it grows. This type of growth is typical of recurrence relationships that have the form U_{n+1}=R U_n\pm k with R>1.
Consider the recurrence relation u_{n+1}=2u_n+2, u_0=2.
Find u_1.
Find u_2.
Find u_3.
Find u_4.
Complete the table of values.
n | 1 | 2 | 3 | 4 |
---|---|---|---|---|
u_n |
Graph the relation.
Find the largest value of n for which u_n<22.
By using a graph, we can understand the behaviour of a recurrence relation whether the values are increasing or decreasing or whether the relation is linear or not.