On Types of Distance Fibonacci Numbers Generated by Number

advertisement
Hindawi Publishing Corporation
Journal of Applied Mathematics
Volume 2014, Article ID 491591, 8 pages
http://dx.doi.org/10.1155/2014/491591
Research Article
On Types of Distance Fibonacci Numbers Generated by
Number Decompositions
Anetta Szynal-Liana, Andrzej WBoch, and Iwona WBoch
Faculty of Mathematics and Applied Physics, Rzeszow University of Technology, Al. Powstańców Warszawy 12, 35-959 Rzeszow, Poland
Correspondence should be addressed to Andrzej WΕ‚och; awloch@prz.edu.pl
Received 2 June 2014; Revised 21 August 2014; Accepted 23 August 2014; Published 23 October 2014
Academic Editor: Ali R. Ashrafi
Copyright © 2014 Anetta Szynal-Liana et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
We introduce new types of distance Fibonacci numbers which are closely related with number decompositions. Using special
decompositions of the number 𝑛 we give a sequence of identities for them. Moreover, we give matrix generators for distance
Fibonacci numbers and their direct formulas.
1. Introduction
The 𝑛th Fibonacci numbers 𝐹𝑛 are defined by recurrence
relation 𝐹𝑛 = 𝐹𝑛−1 +𝐹𝑛−2 , 𝑛 ≥ 2 with the initial conditions 𝐹0 =
𝐹1 = 1. There are many generalizations of the Fibonacci numbers 𝐹𝑛 with respect to one or more parameters; see for example [1–3]. In [1] the distance Fibonacci numbers 𝐹𝑑(1) (π‘˜, 𝑛)
were introduced and studied. We recall this definition.
Let π‘˜ ≥ 2, 𝑛 ≥ 0 be integers. The distance Fibonacci numbers of the first kind 𝐹𝑑(1) (π‘˜, 𝑛) are defined recursively in the
following way:
(1)
𝐹𝑑
(1)
(π‘˜, 𝑛) = 𝐹𝑑
In this section we introduce two kinds of distance Fibonacci
numbers. Some relations between numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛) for 𝑖 =
1, 2, 3 will be studied.
Let π‘˜ ≥ 2, 𝑛 ≥ 0 be integers. We define the 𝑛th distance
Fibonacci numbers of the second kind 𝐹𝑑(2) (π‘˜, 𝑛) by the π‘˜th
order linear recurrence relation of the form
𝐹𝑑(2) (π‘˜, 𝑛) = 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ + 1)
+ 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜)
(π‘˜, 𝑛 − π‘˜ + 1)
+ 𝐹𝑑(1) (π‘˜, 𝑛 − π‘˜)
2. Distance Fibonacci Numbers
𝐹𝑑(2) (π‘˜, 𝑛) and 𝐹𝑑(3) (π‘˜, 𝑛)
for 𝑛 ≥ π‘˜,
(1)
and 𝐹𝑑(1) (π‘˜, 𝑛) = 1 for 𝑛 = 0, . . . , π‘˜ − 1.
We will call the numbers 𝐹𝑑(1) (π‘˜, 𝑛) the distance Fibonacci numbers of the first kind. The number 𝐹𝑑(1) (π‘˜, 𝑛) is closely
related to the special quasi π‘˜-decomposition of the number 𝑛;
see [1].
In this paper we define other three types of distance
Fibonacci numbers which are also related to the special
number decomposition. Moreover we shall show relations
between all three types of distance Fibonacci numbers. Next
we study their matrix generators and direct formulas.
for 𝑛 ≥ π‘˜,
(2)
with the initial conditions
𝐹𝑑(2) (π‘˜, 𝑛) = 0 for 𝑛 = 0, . . . , π‘˜ − 2,
𝐹𝑑(2) (π‘˜, π‘˜ − 1) = 1,
𝐹𝑑(2) (1, 1) = 1,
𝐹𝑑(2) (2, 2) = 2,
𝐹𝑑(2) (π‘˜, π‘˜) = 1, for π‘˜ ≥ 3.
If π‘˜ = 2, 𝑛 ≥ 1, then 𝐹𝑑(2) (π‘˜, 𝑛) gives the Fibonacci numbers 𝐹𝑛 .
2
Journal of Applied Mathematics
Let π‘˜ ≥ 2, 𝑛 ≥ 0 be integers. We define the 𝑛th distance
Fibonacci numbers of the third kind 𝐹𝑑(3) (π‘˜, 𝑛) by the π‘˜th
order linear recurrence relation of the form
(3)
𝐹𝑑
(3)
(π‘˜, 𝑛) = 𝐹𝑑
π‘˜−2
𝜎−π‘Ÿ0 (π‘˜, 𝑛) = ∑ 𝜎−π‘Ÿ (π‘˜, 𝑛) + 𝜎0 (π‘˜, 𝑛) ,
(π‘˜, 𝑛 − π‘˜ + 1)
+ 𝐹𝑑(3) (π‘˜, 𝑛 − π‘˜)
numbers 𝜎−π‘Ÿ (π‘˜, 𝑛), 𝜎−π‘Ÿ0 (π‘˜, 𝑛), and 𝜎0 (π‘˜, 𝑛) (resp., 𝜎+π‘Ÿ (π‘˜, 𝑛),
𝜎+π‘Ÿ0 (π‘˜, 𝑛), and 𝜎0 (π‘˜, 𝑛)):
(5)
π‘Ÿ=1
for 𝑛 ≥ 2π‘˜ − 1,
(3)
π‘˜−2
𝜎+π‘Ÿ0 (π‘˜, 𝑛) = ∑ 𝜎+π‘Ÿ (π‘˜, 𝑛) + 𝜎0 (π‘˜, 𝑛) ,
(6)
π‘Ÿ=1
with the initial conditions
π‘˜−2
𝜎 (π‘˜, 𝑛) =
∑ πœŽπ‘— (π‘˜, 𝑛) .
(7)
𝑗=−(π‘˜−2)
(3)
𝐹𝑑 (π‘˜, 𝑛) = 1 for 𝑛 = 0, ..., π‘˜ − 1,
Theorem 1 (see [1]). Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ − 1 be integers. Then
𝜎+π‘Ÿ0 (π‘˜, 𝑛) = 𝐹𝑑(1) (π‘˜, 𝑛).
(3)
𝐹𝑑 (2, 2) = 2,
for π‘˜ ≥ 3 𝐹𝑑(3) (π‘˜, π‘˜) = 3 = 𝐹𝑑(3) (π‘˜, 2π‘˜ − 2),
For the proof of the next theorem we will need the following lemma.
for π‘˜ + 1 ≤ 𝑛 ≤ 2π‘˜ − 3 𝐹𝑑(3) (π‘˜, 𝑛) = 4.
Lemma 2. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ − 1 be integers. Then
If π‘˜ = 2, then 𝐹𝑑(3) (π‘˜, 𝑛) gives the classical Fibonacci
numbers.
Now we give an interpretation of the numbers 𝐹𝑑(1) (π‘˜, 𝑛),
(2)
𝐹𝑑 (π‘˜, 𝑛), and 𝐹𝑑(3) (π‘˜, 𝑛) with respect to special decompositions of the number 𝑛.
By a decomposition of a number 𝑛, 𝑛 ≥ 1, we mean an
ordered number partition of it. For example for 𝑛 = 3 we
have the following four decompositions: 1 + 1 + 1, 2 + 1, 1+
2, and 3. In this paper we study special decompositions of
a number 𝑛 which are closely related to distance Fibonacci
numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛), for 𝑖 = 1, 2, 3.
Let 1 ≤ π‘Ÿ ≤ π‘˜ − 2 be a fixed integer. A decomposition of
the number 𝑛 ≥ π‘˜ − 1 of the form π‘Ÿ + 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 (resp.,
𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 + π‘Ÿ) where 𝑛𝑖 ∈ {π‘˜, π‘˜ − 1}, 𝑖 = 1, . . . , 𝑝
is called an π‘Ÿ− π‘˜-decomposition (resp., π‘Ÿ+ π‘˜-decomposition).
We denote the number of all π‘Ÿ− π‘˜-decompositions (resp., π‘Ÿ+ π‘˜decompositions) by 𝜎−π‘Ÿ (π‘˜, 𝑛) (resp., 𝜎+π‘Ÿ (π‘˜, 𝑛)). Clearly
𝜎−π‘Ÿ (π‘˜, 𝑛) = 𝜎+π‘Ÿ (π‘˜, 𝑛) .
(4)
A decomposition of the number 𝑛 ≥ π‘˜ − 1 of the form 𝑛1 +
𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 , where 𝑛𝑖 ∈ {π‘˜, π‘˜ − 1}, 𝑖 = 1, . . . , 𝑝, is called a
π‘˜-decomposition. We denote the number of all π‘˜-decompositions by 𝜎0 (π‘˜, 𝑛).
Let 0 ≤ π‘Ÿ0 ≤ π‘˜ − 2 be a fixed integer. A decomposition
of the number 𝑛 ≥ π‘˜ − 1 of the form π‘Ÿ0 + 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ +
𝑛𝑝 (resp., 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 + π‘Ÿ0 ) where 𝑛𝑖 ∈ {π‘˜, π‘˜ − 1},
𝑖 = 1, . . . , 𝑝 is called an π‘Ÿ0− π‘˜-decomposition (resp., π‘Ÿ0+ π‘˜decomposition). Consequently as the above we denote the
number of all π‘Ÿ0− π‘˜-decompositions (resp., π‘Ÿ0+ π‘˜-decompositions) by 𝜎−π‘Ÿ0 (π‘˜, 𝑛) (resp., 𝜎+π‘Ÿ0 (π‘˜, 𝑛)). Clearly for π‘Ÿ0 = 0
a 0− π‘˜-decomposition of 𝑛 is a π‘˜-decomposition and for
π‘Ÿ0 ≥ 1 an π‘Ÿ0− π‘˜-decomposition (resp., π‘Ÿ0+ π‘˜-decompositions)
is an π‘Ÿ− π‘˜-decompositions (resp., π‘Ÿ+ π‘˜-decompositions). From
the above definitions immediately follow relations between
2𝐹𝑑(1) (π‘˜, 𝑛) − 𝐹𝑑(2) (π‘˜, 𝑛) = 𝐹𝑑(3) (π‘˜, 𝑛) ,
for 𝑛 ≥ π‘˜ − 1.
(8)
Proof. If 𝑛 = π‘˜ − 1, then the equality immediately follows.
Assume that the lemma is true for an arbitrary 𝑑 < 𝑛 and
we prove it for 𝑛. Using the definitions of numbers 𝐹𝑑(1) (π‘˜, 𝑛)
and 𝐹𝑑(2) (π‘˜, 𝑛) we obtain that 2𝐹𝑑(1) (π‘˜, 𝑛) − 𝐹𝑑(2) (π‘˜, 𝑛) =
2𝐹𝑑(1) (π‘˜, 𝑛−π‘˜)+2𝐹𝑑(1) (π‘˜, 𝑛−π‘˜+1)−𝐹𝑑(2) (π‘˜, 𝑛−π‘˜)−𝐹𝑑(2) (π‘˜, 𝑛−
π‘˜ + 1) = 𝐹𝑑(3) (π‘˜, 𝑛) by the induction’s hypothesis.
We can write the above lemma also in the following form.
Corollary 3. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ − 1 be integers. Then
𝐹𝑑(1) (π‘˜, 𝑛) − 𝐹𝑑(2) (π‘˜, 𝑛) = 𝐹𝑑(3) (π‘˜, 𝑛) − 𝐹𝑑(1) (π‘˜, 𝑛) .
(9)
Theorem 4. Let π‘˜ ≥ 2, 𝑛 ≥ 1, 1 ≤ π‘Ÿ ≤ π‘˜ − 2 be integers. Then
(i) 𝜎−π‘Ÿ0 (π‘˜, 𝑛) = 𝜎+π‘Ÿ0 (π‘˜, 𝑛) = 𝐹𝑑(1) (π‘˜, 𝑛),
(ii) 𝜎0 (π‘˜, 𝑛) = 𝐹𝑑(2) (π‘˜, 𝑛),
(iii) 𝜎(π‘˜, 𝑛) = 𝐹𝑑(3) (π‘˜, 𝑛).
Proof. The equality (i) follows immediately by Theorem 1 and
(4).
We shall show that 𝜎0 (π‘˜, 𝑛) = 𝐹𝑑(2) (π‘˜, 𝑛). If 𝑛 = 1, 2, . . . ,
π‘˜ − 2, then there is no π‘˜-decomposition of the number 𝑛 into
parts π‘˜ − 1 and π‘˜. So 𝜎0 (π‘˜, 𝑛) = 0 = 𝐹𝑑(2) (π‘˜, 𝑛). If 𝑛 = π‘˜ −
1, π‘˜, then there is a unique π‘˜-decomposition of the number
𝑛; hence 𝜎0 (π‘˜, 𝑛) = 1 = 𝐹𝑑(2) (π‘˜, 𝑛). Let 𝑛 ≥ π‘˜ + 1. Assume
that the equality holds for an arbitrary 𝑑 < 𝑛. We shall show
that 𝜎0 (π‘˜, 𝑛) = 𝐹𝑑(2) (π‘˜, 𝑛). Let 𝑛 = 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 be a
π‘˜-decomposition of the number 𝑛 into parts π‘˜ and π‘˜ − 1. If
𝑛𝑝 = π‘˜, then 𝑛 = 𝑛1 + 𝑛2 + . . . + 𝑛𝑝−1 + π‘˜ so 𝑛 − π‘˜ = 𝑛1 +
𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝−1 . By induction’s hypothesis there are 𝐹𝑑(2) (π‘˜, 𝑛 −
π‘˜) π‘˜-decompositions in this case. If 𝑛𝑝 = π‘˜ − 1 then proving
analogously we obtain 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ + 1) π‘˜-decompositions
Journal of Applied Mathematics
3
of the form 𝑛 = 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝−1 + π‘˜ − 1. From the above
we have 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜) + 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ + 1) π‘˜-decompositions
of the number 𝑛 into parts π‘˜ and π‘˜ − 1, and by the definition
of 𝐹𝑑(2) (π‘˜, 𝑛) it follows that 𝜎0 (π‘˜, 𝑛) = 𝐹𝑑(2) (π‘˜, 𝑛).
Now we shall prove that 𝜎(π‘˜, 𝑛) = 𝐹𝑑(3) (π‘˜, 𝑛). From the
definition of 𝜎(π‘˜, 𝑛) we obtain that
𝜎 (π‘˜, 𝑛) =
Theorem 7. Let π‘˜ ≥ 3, 𝑛 ≥ π‘˜ be integers. Then
π‘˜−2 π‘˜−2
∑ ∑ 𝐹𝑑(2) (π‘˜, 𝑛 − (𝑖 + 𝑗))
𝑖=1 𝑗=1
(13)
π‘˜−2
= ∑ (𝐹𝑑(1) (π‘˜, 𝑛 − 𝑖) − 𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖)) ,
π‘˜−2
𝑖=1
∑ πœŽπ‘— (π‘˜, 𝑛)
for natural 𝑝 and 𝑖 = 1, 2, . . . , π‘˜ − (𝑝 + 2).
𝑗=−(π‘˜−2)
π‘˜−2
= 2 ∑ 𝜎+π‘Ÿ (π‘˜, 𝑛) + 𝜎0 (π‘˜, 𝑛)
π‘Ÿ=1
(10)
= 𝐹𝑑(1) (π‘˜, 𝑛)
π‘˜−2
+ ∑ 𝜎+π‘Ÿ (π‘˜, 𝑛) + 𝜎0 (π‘˜, 𝑛) − 𝜎0 (π‘˜, 𝑛)
π‘Ÿ=1
Proof. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integers. Let 𝑛 = 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝
be a decomposition πœ‚ of the number 𝑛, where 𝑛𝑖 ∈ {π‘˜, π‘˜ −
1}, for 𝑖 = 2, . . . , 𝑝 − 1 and 𝑛1 , 𝑛𝑝 ∈ {1, . . . , π‘˜ − 2}. Then the
number of such decomposition is equal to the number of π‘˜decomposition of the number 𝑛 − (𝑛1 + 𝑛𝑝 ). By Theorem 4(ii)
we obtain that we have 𝐹𝑑(2) (π‘˜, 𝑛 − (𝑛1 + 𝑛𝑝 )) decompositions
πœ‚. Since 𝑛1 , 𝑛𝑝 ∈ {1, . . . , π‘˜ − 2} it is clear that totally we have
= 2𝐹𝑑(1) (π‘˜, 𝑛) − 𝐹𝑑(2) (π‘˜, 𝑛) ,
π‘˜−2 π‘˜−2
by the statements (i) and (ii) of this theorem.
Then the statement (iii) follows immediately by Lemma 2,
which ends the proof.
𝑖=1 𝑗=1
∑ ∑ 𝐹𝑑(2) (π‘˜, 𝑛 − (𝑖 + 𝑗))
Theorem 5. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ − 1 and 1 ≤ π‘Ÿ ≤ π‘˜ − 2 be integers.
Then
𝜎−π‘Ÿ (π‘˜, π‘π‘˜ + 𝑖) = 2𝑝 ,
decompositions of the number 𝑛.
On the other hand, we have that the total number of
decompositions πœ‚ is equal to the number of π‘Ÿ− π‘˜-decompositions of the number 𝑛−𝑛𝑝 . Since 𝑛𝑝 ∈ {1, . . . , π‘˜−2} by formula
(5) and Theorem 4 we obtain
(11)
π‘˜−2
∑ (𝐹𝑑(1) (π‘˜, 𝑛 − 𝑖) − 𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖)) .
for natural 𝑝 and 𝑖 = 1, 2, . . . , π‘˜ − (𝑝 + 2).
From the above it immediately follows that
π‘˜−2 π‘˜−2
∑ ∑ 𝐹𝑑(2) (𝑛 − (𝑖 + 𝑗))
𝑖=1 𝑗=1
(16)
π‘˜−2
Now we give applications of distance Fibonacci numbers
for counting of the number of other special decompositions
of the number 𝑛.
Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integers and let 𝜎±π‘Ÿ0 (π‘˜, 𝑛) be the numbers of all decomposition of the number 𝑛 = 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 ,
where 𝑛𝑖 ∈ {π‘˜, π‘˜−1}, for 𝑖 = 2, . . . , 𝑝−1 and 𝑛1 , 𝑛𝑝 ∈ {1, . . . , π‘˜}.
Theorem 6. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integers. Then
(1)
𝜎±π‘Ÿ0 (π‘˜, 𝑛) = ∑𝐹𝑑
(π‘˜, 𝑛 − π‘Ÿ) .
(15)
𝑖=1
Proof. Let 𝑝 be natural and 𝑖 = 1, 2, . . . , π‘˜−(𝑝+2). The number
π‘π‘˜ + 𝑖 is equal to (𝑖 + 𝑙) + 𝑙 ⋅ (π‘˜ − 1) + (𝑝 − 𝑙) ⋅ π‘˜, for 𝑙 =
0, 1, . . . , 𝑝, and all π‘Ÿ− π‘˜-decompositions of π‘π‘˜+𝑖 have the form
(𝑖 + 𝑙) + 𝑛1 + . . . + 𝑛𝑝 . We can put π‘˜ − 1 on 𝑙 positions, so we
𝑝
have ( 𝑝𝑙 ) possibilities. The sum ∑𝑙=0 ( 𝑝𝑙 ) is equal to 2𝑝 which
ends the proof.
π‘˜
(14)
(12)
π‘Ÿ=1
Proof. Let 𝑛 = 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝 be a decomposition of the
number 𝑛, where 𝑛𝑖 ∈ {π‘˜, π‘˜−1}, for 𝑖 = 2, . . . , 𝑝−1 and 𝑛1 , 𝑛𝑝 ∈
{1, . . . , π‘˜}. Then 𝑛 − 𝑛𝑝 = 𝑛1 + 𝑛2 + ⋅ ⋅ ⋅ + 𝑛𝑝−1 , where 𝑛1 + 𝑛2 +
⋅ ⋅ ⋅ + 𝑛𝑝−1 is either a π‘Ÿ− π‘˜-decomposition or a π‘˜-decomposition
of the number 𝑛 − 𝑛𝑝 . Since 𝑛𝑝 ∈ {1, . . . , π‘˜} by Theorem 4(i) it
follows that 𝜎±π‘Ÿ0 (π‘˜, 𝑛) = ∑π‘˜π‘Ÿ=1 𝐹𝑑(1) (π‘˜, 𝑛 − π‘Ÿ), which ends the
proof.
= ∑ (𝐹𝑑(1) (π‘˜, 𝑛 − 𝑖) − 𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖)) ,
𝑖=1
which ends the proof.
3. Identities for 𝐹𝑑(1) (π‘˜, 𝑛),
𝐹𝑑(2) (π‘˜, 𝑛), and 𝐹𝑑(3) (π‘˜, 𝑛)
In this section we give some identities for distance Fibonacci
numbers: 𝐹𝑑(𝑖) (π‘˜, 𝑛), for 𝑖 = 1, 2, 3.
Theorem 8. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integers. Then
3
∑π‘Žπ‘— 𝐹𝑑(𝑗) (π‘˜, 𝑛)
𝑗=1
𝑝
3
𝑖=0
𝑗=1
𝑝
= ∑ ( ) ( ∑π‘Žπ‘— 𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖)) .
𝑖
(17)
4
Journal of Applied Mathematics
Proof (by induction on 𝑛). For 𝑛 = π‘˜ we have 𝑝 = 1 and
Proof. (iv) This formula directly follows from Theorem 8 putting π‘Ž1 = 1 and π‘Ž2 = π‘Ž3 = 0. For (v) and (vi) analogously.
3
∑ π‘Žπ‘— 𝐹𝑑(𝑗) (π‘˜, π‘˜)
Theorem 10. Let π‘˜ ≥ 3, 𝑛 ≥ π‘˜ be integers. Then
𝑗=1
π‘˜−2
3
(𝑗)
= ∑ π‘Žπ‘— (𝐹𝑑
(𝑗)
(π‘˜, π‘˜ − π‘˜) + 𝐹𝑑
𝐹𝑑(1) (π‘˜, 𝑛) = ∑ 𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖) .
(π‘˜, π‘˜ − π‘˜ + 1))
𝑗=1
3
(𝑗)
= ∑ π‘Žπ‘— (𝐹𝑑
(𝑗)
(π‘˜, 0) + 𝐹𝑑
(π‘˜, 1))
𝑗=1
(18)
3
= ∑ π‘Žπ‘— (𝐹𝑑(𝑗) (π‘˜, π‘˜ − 1 ⋅ π‘˜ + 0)
𝑗=1
+𝐹𝑑(𝑗) (π‘˜, π‘˜ − 1 ⋅ π‘˜ + 1))
1
3
𝑖=0
𝑗=1
(20)
𝑖=0
1
= ∑ ( ) ∑π‘Žπ‘— 𝐹𝑑(𝑗) (π‘˜, 𝑖) .
𝑖
Proof (by induction on 𝑛). For 𝑛 = π‘˜ we have 𝐹𝑑(1) (π‘˜, π‘˜) =
𝐹𝑑(1) (π‘˜, 0) + 𝐹𝑑(1) (π‘˜, 1) = 1 + 1 = 2. The right side of (20) has
(2)
(2)
(2)
the form ∑π‘˜−2
𝑖=0 𝐹𝑑 (π‘˜, π‘˜ − 𝑖) = 𝐹𝑑 (π‘˜, π‘˜ − 0) + 𝐹𝑑 (π‘˜, π‘˜ −
(2)
(2)
(2)
1) + 𝐹𝑑 (π‘˜, π‘˜ − 2) + ⋅ ⋅ ⋅ + 𝐹𝑑 (π‘˜, π‘˜ − (π‘˜ − 2)) = 𝐹𝑑 (π‘˜, π‘˜) +
𝐹𝑑(2) (π‘˜, π‘˜ − 1) + 𝐹𝑑(2) (π‘˜, π‘˜ − 2) + ⋅ ⋅ ⋅ + 𝐹𝑑(2) (π‘˜, 2) = 1 + 1 + 0 +
⋅ ⋅ ⋅ + 0 = 2.
Let 𝑛 > π‘˜. Assume that (20) is true for arbitrary 𝑑 < 𝑛 and
we prove it for 𝑛. Using induction’s assumption for 𝑑 = 𝑛 − π‘˜
and 𝑑 = 𝑛 − π‘˜ + 1 we have
𝐹𝑑(1) (π‘˜, 𝑛)
Let 𝑛 > π‘˜. Assume that (17) is true for arbitrary 𝑑 < 𝑛 and
we prove it for 𝑛. Using induction’s assumption for 𝑑 = 𝑛 − π‘˜
and 𝑑 = 𝑛 − π‘˜ + 1 we have
= 𝐹𝑑(1) (π‘˜, 𝑛 − π‘˜) + 𝐹𝑑(1) (π‘˜, 𝑛 − π‘˜ + 1)
π‘˜−2
= ∑ 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ − 𝑖)
𝑖=0
𝐹𝑑(𝑗) (π‘˜, 𝑛)
π‘˜−2
+ ∑ 𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ + 1 − 𝑖)
= 𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘˜) + 𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘˜ + 1)
𝑖=0
𝑝
𝑝
= ∑ ( ) 𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘˜ − π‘π‘˜ + 𝑖)
𝑖
π‘˜−2
= ∑ [𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ − 𝑖)
𝑖=0
𝑖=0
𝑝
𝑝
+ ∑ ( ) 𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘˜ + 1 − π‘π‘˜ + 𝑖)
𝑖
+𝐹𝑑(2) (π‘˜, 𝑛 − π‘˜ + 1 − 𝑖)]
𝑖=0
π‘˜−2
𝑝
𝑝
= ∑ ( ) [𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘˜ − π‘π‘˜ + 𝑖)
𝑖
= ∑ [𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖 − π‘˜)
𝑖=0
𝑖=0
+𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖 − π‘˜ + 1)]
+𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘˜ + 1 − π‘π‘˜ + 𝑖)]
𝑝
π‘˜−2
𝑝
= ∑ ( ) [𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖 − π‘˜)
𝑖
= ∑ 𝐹𝑑(2) (π‘˜, 𝑛 − 𝑖) ,
𝑖=0
𝑖=0
+𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖 − π‘˜ + 1)]
which ends the proof.
𝑝
𝑝
= ∑ ( ) 𝐹𝑑(𝑗) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖) ,
𝑖
𝑖=0
(19)
which ends the proof.
Corollary 9. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integers. Then
𝑝
(iv) 𝐹𝑑(1) (π‘˜, 𝑛) = ∑𝑖=0 ( 𝑝𝑖 ) 𝐹𝑑(1) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖),
𝑝
(v) 𝐹𝑑(2) (π‘˜, 𝑛) = ∑𝑖=0 ( 𝑝𝑖 ) 𝐹𝑑(2) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖),
𝑝
(vi) 𝐹𝑑(3) (π‘˜, 𝑛) = ∑𝑖=0 ( 𝑝𝑖 ) 𝐹𝑑(3) (π‘˜, 𝑛 − π‘π‘˜ + 𝑖).
(21)
4. Matrix Generators and
Combinatorial Formulas for 𝐹𝑑(1) (π‘˜, 𝑛),
𝐹𝑑(2) (π‘˜, 𝑛), and 𝐹𝑑(3) (π‘˜, 𝑛)
Matrix methods are important in recurrence relations. In the
last decades some mathematicians have studied to find miscellaneous affinities between matrices and linear recurrences.
Using matrix methods different identities and algebraic
representations of considered sequences can be obtained, for
instance [1, 2, 4–6]. Theory of Fibonacci numbers was previously complemented by the theory of so-called the Fibonacci
𝑄-matrix or the golden matrix. It is worth mentioning that
Journal of Applied Mathematics
5
the American mathematician V. Hoggat was one of the first
mathematicians who paid the attention to the 𝑄-matrix.
For the classical Fibonacci sequence the matrix generators,
named as the golden matrix, have the form 𝑄 = [ 11 10 ] and
𝐹
𝐹𝑛
], for 𝑛 ≥ 1.
𝑄𝑛 = [ 𝐹𝑛+1
𝑛 𝐹𝑛−1
Golden number and golden section have many interesting applications in different areas of science (physics,
chemistry, and mechanics); see for example [7, 8]. In this
section we give the matrix generators for distance Fibonacci
numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛), where 𝑖 = 1, 2, 3. The matrix generator
of the distance Fibonacci numbers of the first kind 𝐹𝑑(1) (π‘˜, 𝑛)
was introduced in [1] and in this paper we apply this method
for all kinds of distance Fibonacci numbers.
Let π‘˜ ≥ 2 be a fixed integer. Let π‘„π‘˜ = [π‘žπ‘‘π‘— ] be a square
matrix of size π‘˜. For a fixed 1 ≤ 𝑑 ≤ π‘˜ an element π‘žπ‘‘1 is equal
to the coefficient of 𝐹𝑑(𝑖) (π‘˜, 𝑛 − 𝑑) in the recurrence formula
for the distance Fibonacci numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 = 1, 2, 3. For
𝑗 ≥ 2 we define π‘žπ‘‘π‘— as follows:
{1 if 𝑗 = 𝑑 + 1
π‘žπ‘‘π‘— = {
0 otherwise.
{
𝐴(𝑖)
π‘˜
𝐹𝑑(𝑖) (π‘˜, 2π‘˜ − 2 + 𝑒 (𝑖))
[
[𝐹𝑑(𝑖) (π‘˜, 2π‘˜ − 3 + 𝑒 (𝑖))
[
[
[
..
=[
.
[
[
[ 𝐹𝑑(𝑖) (π‘˜, π‘˜ + 𝑒 (𝑖))
[
(𝑖)
[ 𝐹𝑑 (π‘˜, π‘˜ − 1 + 𝑒 (𝑖))
(22)
for 𝑖 = 1, 2
for 𝑖 = 3.
𝑄2 = [
1 1
],
1 0
0
[
[0
0 1 0
[
[.
[1 0 1]
𝑄3 = [
] , . . . , π‘„π‘˜ = [
[ ..
[
[1
[1 0 0]
[
1 0 ⋅⋅⋅ 0
]
0 1 ⋅ ⋅ ⋅ 0]
]
.. ..
.. ]
]
. .
.] ,
]
0 0 ⋅ ⋅ ⋅ 1]
]
(25)
and the matrix π‘„π‘˜ is named as the distance Fibonacci
matrix or the generator of the distance Fibonacci numbers
𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 = 1, 2, 3.
For a fixed 1 ≤ 𝑖 ≤ 3 we define the square matrix 𝐴(𝑖)
π‘˜ of
size π‘˜ named as the matrix of initial conditions of the form
(24)
Theorem 11. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integer. Then for a fixed
1 ≤ 𝑖 ≤ 3 holds
𝐹𝑑(𝑖) (π‘˜, 𝑛 + 2π‘˜ − 2 + 𝑒 (𝑖)) 𝐹𝑑(𝑖) (π‘˜, 𝑛 + 2π‘˜ − 3 + 𝑒 (𝑖))
[ (𝑖)
[𝐹𝑑 (π‘˜, 𝑛 + 2π‘˜ − 3 + 𝑒 (𝑖)) 𝐹𝑑(𝑖) (π‘˜, 𝑛 + 2π‘˜ − 4 + 𝑒 (𝑖))
[
[
𝑛
..
..
[
𝐴(𝑖)
𝑄
=
.
.
π‘˜ π‘˜
[
[
(𝑖)
[ 𝐹𝑑(𝑖) (π‘˜, 𝑛 + π‘˜ + 𝑒 (𝑖))
𝐹𝑑 (π‘˜, 𝑛 + π‘˜ − 1 + 𝑒 (𝑖))
[
(𝑖)
𝐹𝑑(𝑖) (π‘˜, 𝑛 + π‘˜ + 𝑒 (𝑖))
[ 𝐹𝑑 (π‘˜, 𝑛 + π‘˜ − 1 + 𝑒 (𝑖))
Since the proof is analogous as in [1] we omit it. To obtain
other matrix generators apart distance Fibonacci numbers
𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 = 1, 2, 3, we define a collection of special sequences which are given by the same π‘˜th order linear recurrence
relations as 𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 = 1, 2, 3. These sequences give
auxiliary tools for other matrix generators of 𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 =
1, 2, 3 and their explicit formulas.
(23)
[1 0 0 ⋅ ⋅ ⋅ 0]
𝐹𝑑(𝑖) (π‘˜, 2π‘˜ − 3 + 𝑒 (𝑖)) ⋅ ⋅ ⋅ 𝐹𝑑(𝑖) (π‘˜, π‘˜ − 1 + 𝑒 (𝑖))
]
𝐹𝑑(𝑖) (π‘˜, 2π‘˜ − 4 + 𝑒 (𝑖)) ⋅ ⋅ ⋅ 𝐹𝑑(𝑖) (π‘˜, π‘˜ − 2 + 𝑒 (𝑖))]
]
]
]
..
..
],
.
d
.
]
]
(𝑖)
(𝑖)
𝐹𝑑 (π‘˜, π‘˜ − 1 + 𝑒 (𝑖)) ⋅ ⋅ ⋅ 𝐹𝑑 (π‘˜, 1 + 𝑒 (𝑖)) ]
]
𝐹𝑑(𝑖) (π‘˜, π‘˜ − 2 + 𝑒 (𝑖)) ⋅ ⋅ ⋅ 𝐹𝑑(𝑖) (π‘˜, 0 + 𝑒 (𝑖)) ]
where
𝑖−1
𝑒 (𝑖) = {
π‘˜−1
In other words,
⋅ ⋅ ⋅ 𝐹𝑑(𝑖) (π‘˜, 𝑛 + π‘˜ − 1 + 𝑒 (𝑖))
]
⋅ ⋅ ⋅ 𝐹𝑑(𝑖) (π‘˜, 𝑛 + π‘˜ − 2 + 𝑒 (𝑖))]
]
]
..
].
d
.
]
]
(𝑖)
⋅ ⋅ ⋅ 𝐹𝑑 (π‘˜, 𝑛 + 1 + 𝑒 (𝑖)) ]
]
⋅ ⋅ ⋅ 𝐹𝑑(𝑖) (π‘˜, 𝑛 + 0 + 𝑒 (𝑖)) ]
(26)
Let π‘˜ ≥ 2, 𝑛 ≥ 0 be integers. Let Fπ‘˜ = {𝐹𝑑𝑙(4) (π‘˜, 𝑛), 𝑙 = 0,
1, . . . , π‘˜ − 1}, where {𝐹𝑑𝑙(4) (π‘˜, 𝑛)} is the sequence defined as
follows:
𝐹𝑑𝑙(4) (π‘˜, 𝑛) = 𝐹𝑑𝑙(4) (π‘˜, 𝑛 − π‘˜ + 1) + 𝐹𝑑𝑙(4) (π‘˜, 𝑛 − π‘˜) ,
for 𝑛 ≥ π‘˜,
(27)
6
Journal of Applied Mathematics
with the initial conditions 𝐹𝑑𝑙(4) (π‘˜, 𝑙) = 1 and 𝐹𝑑𝑙(4) (π‘˜, 𝑗) = 0
for 𝑗 =ΜΈ 𝑙.
The number 𝐹𝑑1(4) (π‘˜, 𝑛) will be also denoted shortly by
(4)
𝐹𝑑 (π‘˜, 𝑛) and named as the 𝑛th distance Fibonacci number
of the fourth kind. If π‘˜ = 2 and 𝑛 ≥ 1, then 𝐹𝑑(4) (2, 𝑛) = 𝐹𝑛−1 .
By simple observation we obtain the following relations
between numbers 𝐹𝑑𝑙(4) (π‘˜, 𝑛), for 𝑙 = 0, 1, . . . , π‘˜ − 1:
𝐹𝑑(1) (π‘˜, 𝑙) if 𝑛 = 𝑙
𝐹𝑑𝑙(1) (π‘˜, 𝑛) = {
0
if 𝑛 =ΜΈ 𝑙
Then
for 2 ≤ 𝑙 ≤ π‘˜ − 1.
(28)
(1)
= 𝐹𝑑0(1) (π‘˜, 𝑛) + ⋅ ⋅ ⋅ + πΉπ‘‘π‘˜−1
(π‘˜, 𝑛)
= 𝐹𝑑(1) (π‘˜, 0) 𝐹𝑑0(1) (π‘˜, 𝑛)
Using sequences from the collection Fπ‘˜ we can generate
the distance Fibonacci numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 = 1, 2, 3:
(1)
𝐹𝑑
π‘˜−1
(1)
(π‘˜, 𝑛) = ∑ 𝐹𝑑
𝑙=0
(π‘˜, 𝑙) 𝐹𝑑𝑙(4)
(π‘˜, 𝑛) =
π‘˜−1
∑ 𝐹𝑑𝑙(4)
𝑙=0
(1)
+ ⋅ ⋅ ⋅ + 𝐹𝑑(1) (π‘˜, −1) πΉπ‘‘π‘˜−1
(π‘˜, 𝑛)
(π‘˜, 𝑛) ,
π‘˜−1
= ∑ 𝐹𝑑(4) (π‘˜, 𝑛 − 𝑑) .
π‘˜
𝑑=0
𝑙=π‘˜−1
= ∑
𝑙=π‘˜−1
(31)
(1)
= 1𝐹𝑑0(1) (π‘˜, 𝑛) + ⋅ ⋅ ⋅ + 1πΉπ‘‘π‘˜−1
(π‘˜, 𝑛)
(4)
𝐹𝑑(2) (π‘˜, 𝑛) = ∑ 𝐹𝑑(2) (π‘˜, 𝑙) 𝐹𝑑𝑙−1
(π‘˜, 𝑛 − 1)
π‘˜
for 𝑛 ≤ π‘˜ − 1. (30)
𝐹𝑑𝑙(1) (π‘˜, 𝑛)
𝐹𝑑0(4) (π‘˜, 𝑛) = 𝐹𝑑(4) (π‘˜, 𝑛 − π‘˜ + 1) ,
𝐹𝑑𝑙(4) (π‘˜, 𝑛) = 𝐹𝑑(4) (π‘˜, 𝑛 − 𝑙 + 1)
Proof. Let 𝐹𝑑𝑙(1) (π‘˜, 𝑛) denote a sequence defined by the same
recurrence as 𝐹𝑑(1) (π‘˜, 𝑛) with initial conditions
We prove analogously formulas (ii) and (iii).
(4)
𝐹𝑑𝑙−1
(π‘˜, 𝑛 − 1)
for 𝑛 ≥ 1,
Using the above theorem we obtain a new matrix generator for distance Fibonacci numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛), 𝑖 = 1, 2, 3.
𝐹𝑑(3) (π‘˜, 𝑛)
Corollary 13. For the distance Fibonacci numbers 𝐹𝑑(𝑖) (π‘˜, 𝑛),
𝑖 = 1, 2, 3, Theorem 12 gives its matrix generator, respectively:
π‘˜−1
= ∑ 𝐹𝑑(3) (π‘˜, π‘˜ − 1 + 𝑙) 𝐹𝑑𝑙(4) (π‘˜, 𝑛 − (π‘˜ − 1))
𝑛−𝑑
(1) ∑π‘˜−1
𝑑=0 π‘„π‘˜ for 𝑛 ≥ π‘˜,
𝑙=0
for 𝑛 ≥ π‘˜ − 1.
(29)
Theorem 12. Let π‘˜ ≥ 2 be integer. Then
(4)
(i) 𝐹𝑑(1) (π‘˜, 𝑛) = ∑π‘˜−1
𝑑=0 𝐹𝑑 (π‘˜, 𝑛 − 𝑑) for 𝑛 ≥ π‘˜,
(2) π‘„π‘˜π‘›−(π‘˜−2) + π‘„π‘˜π‘›−(π‘˜−1) for 𝑛 ≥ π‘˜,
𝑛−(π‘˜−1)−𝑑+1
+π‘„π‘˜π‘›−2(π‘˜−1)+1
(3) π‘„π‘˜π‘›−2(π‘˜−1) +3π‘„π‘˜π‘›−(π‘˜−1) +4 ∑π‘˜−2
𝑑=2 π‘„π‘˜
for 𝑛 ≥ 2π‘˜ − 1.
Theorem 14. Let π‘˜ ≥ 2, 𝑛 ≥ π‘˜ be integer. Then
𝐹𝑑(4) (π‘˜, 𝑛 + 1) =
(ii) 𝐹𝑑(2) (π‘˜, 𝑛) = 𝐹𝑑(4) (π‘˜, 𝑛 − (π‘˜ − 2)) + 𝐹𝑑(4) (π‘˜, 𝑛 − (π‘˜ − 1))
for 𝑛 ≥ π‘˜,
(iii) 𝐹𝑑(3) (π‘˜, 𝑛) = 𝐹𝑑(4) (π‘˜, 𝑛 − 2(π‘˜ − 1)) + 3𝐹𝑑(4) (π‘˜, 𝑛 − (π‘˜ −
(4)
(4)
1)) + 4 ∑π‘˜−2
𝑑=2 𝐹𝑑 (π‘˜, 𝑛 − (π‘˜ − 1) − 𝑑 + 1) + 3𝐹𝑑 (π‘˜, 𝑛 −
2(π‘˜ − 1) + 1) for 𝑛 ≥ 2π‘˜ − 1.
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛
π‘˜ + π‘˜2
( 1
).
π‘˜1
(32)
Proof. We consider a digraph π·π‘˜ represented by adjacency
matrix π‘„π‘˜ auxiliary (Figure 1).
Note that matrix π‘„π‘˜π‘› has the following form:
𝐹𝑑(4) (π‘˜, 𝑛)
𝐹𝑑(4) (π‘˜, 𝑛 + 1)
[
[ 𝐹𝑑(4) (π‘˜, 𝑛 + 2)
𝐹𝑑(4) (π‘˜, 𝑛 + 1)
[
[
..
..
[
.
.
[
[
(4)
(4)
[𝐹𝑑 (π‘˜, 𝑛 + (π‘˜ − 1)) 𝐹𝑑 (π‘˜, 𝑛 + (π‘˜ − 2))
[
𝐹𝑑(4) (π‘˜, 𝑛 − 1)
𝐹𝑑(4) (π‘˜, 𝑛)
[
It is well known that π‘žπ‘–π‘— ∈ π‘„π‘˜π‘› is equal to the number of all
distinct paths of length 𝑛 between vertices V𝑖 and V𝑗 in the
digraph π·π‘˜ .
∑
⋅ ⋅ ⋅ 𝐹𝑑(4) (π‘˜, 𝑛 − (π‘˜ − 2))
]
⋅ ⋅ ⋅ 𝐹𝑑(4) (π‘˜, 𝑛 − (π‘˜ − 1))]
]
]
..
].
d
.
]
]
]
⋅⋅⋅
𝐹𝑑(4) (π‘˜, 𝑛)
]
(4)
⋅ ⋅ ⋅ 𝐹𝑑 (π‘˜, 𝑛 − (π‘˜ − 1))]
(33)
Each path 𝑃 from V1 to V1 in digraph π·π‘˜ has the following
form: 𝑃 : V1 [ Vπ‘₯1 → V1 [ Vπ‘₯2 → V1 ⋅ ⋅ ⋅ V1 [ Vπ‘₯𝑖 → V1 where
V1 [ Vπ‘₯ denotes the unique shortest path from V1 to Vπ‘₯ in
Journal of Applied Mathematics
7
Table 1: The 𝑛th distance Fibonacci numbers 𝐹𝑑(1) (π‘˜, 𝑛).
𝑛
𝐹𝑑(1) (2, 𝑛)
𝐹𝑑(1) (3, 𝑛)
𝐹𝑑(1) (4, 𝑛)
𝐹𝑑(1) (5, 𝑛)
𝐹𝑑(1) (6, 𝑛)
𝐹𝑑(1) (7, 𝑛)
𝐹𝑑(1) (8, 𝑛)
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
1
1
1
1
1
1
3
3
2
1
1
1
1
1
4
5
2
2
1
1
1
1
5
8
3
2
2
1
1
1
6
13
4
2
2
2
1
1
7
21
5
3
2
2
2
1
8
34
7
4
2
2
2
2
9
55
9
4
3
2
2
2
10
89
12
5
4
2
2
2
11
144
16
7
4
3
2
2
12
233
21
8
4
4
2
2
13
377
28
9
5
4
3
2
14
610
37
12
7
4
4
2
11
144
9
3
0
2
0
0
12
233
12
2
1
1
1
0
13
377
16
4
3
0
2
0
14
610
21
6
3
0
1
1
Table 2: The 𝑛th distance Fibonacci numbers 𝐹𝑑(2) (π‘˜, 𝑛).
𝑛
𝐹𝑑(2) (2, 𝑛)
𝐹𝑑(2) (3, 𝑛)
𝐹𝑑(2) (4, 𝑛)
𝐹𝑑(2) (5, 𝑛)
𝐹𝑑(2) (6, 𝑛)
𝐹𝑑(2) (7, 𝑛)
𝐹𝑑(2) (8, 𝑛)
0
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
2
2
1
0
0
0
0
0
3
3
1
1
0
0
0
0
4
5
1
1
1
0
0
0
5
8
2
0
1
1
0
0
6
13
2
1
0
1
1
0
7
21
3
2
0
0
1
1
8
34
4
1
1
0
0
1
9
55
5
1
2
0
0
0
10
89
7
3
1
1
0
0
𝐹𝑑(2) (π‘˜, 𝑛)
π‘˜−1
1
2
···
3
k−1
k
= ∑
∑
𝑖=π‘˜−2
Figure 1: A digraph π·π‘˜ .
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛−𝑖−1
π‘˜ + π‘˜2
( 1
),
π‘˜1
𝑛 ≥ π‘˜,
𝐹𝑑(3) (π‘˜, 𝑛)
digraph π·π‘˜ . Parts V1 [ Vπ‘₯ → V1 are cycles of the length π‘₯,
where π‘₯ is π‘˜ or π‘˜ − 1. Thus the length of the path 𝑃 is equal to
π‘₯1 + π‘₯2 + ⋅ ⋅ ⋅ + π‘₯𝑖 . There exists one-to-one correspondence
between the path 𝑃 and a tuple (π‘₯1 , π‘₯2 , . . . , π‘₯𝑖 ). If the path
𝑃 has a length 𝑛, then the corresponding tuple is a decomposition of an integer 𝑛 when π‘˜ occurs π‘˜1 times and π‘˜ − 1
occurs π‘˜2 times and π‘˜1 ⋅ π‘˜ + π‘˜2 ⋅ (π‘˜ − 1) = 𝑛. The number
2
). We
of such tuples is equal to the binomial coefficient ( π‘˜1π‘˜+π‘˜
1
determine analogously the number of different paths between
other pairs of vertices.
=
∑
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛−1−2(π‘˜−1)
+3
∑
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛−1−(π‘˜−1)
π‘˜−2
+ 4∑
𝑖=2
+
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛−𝑖−(π‘˜−1)
π‘˜−1
=∑
𝑖=0
∑
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛−2(π‘˜−1)
𝐹𝑑(1) (π‘˜, 𝑛)
π‘˜ + π‘˜2
( 1
)
π‘˜1
∑
Using Theorems 14 and 12 we can prove the following.
Theorem 15. Let π‘˜ ≥ 2 be integer. Then
π‘˜ + π‘˜2
( 1
)
π‘˜1
π‘˜ + π‘˜2
( 1
)
π‘˜1
π‘˜ + π‘˜2
( 1
),
π‘˜1
𝑛 ≥ 2π‘˜ − 1.
∑
π‘˜1 ,π‘˜2
π‘˜1 ⋅π‘˜+π‘˜2 ⋅(π‘˜−1)=𝑛−𝑖−1
π‘˜ + π‘˜2
( 1
),
π‘˜1
𝑛 ≥ π‘˜,
(34)
Tables 1, 2, 3, and 4 present first words of four types of distance Fibonacci numbers 𝐹𝑑(1) (π‘˜, 𝑛), 𝐹𝑑(2) (π‘˜, 𝑛), 𝐹𝑑(3) (π‘˜, 𝑛),
and 𝐹𝑑(4) (π‘˜, 𝑛).
8
Journal of Applied Mathematics
Table 3: The 𝑛th distance Fibonacci numbers 𝐹𝑑(3) (π‘˜, 𝑛).
𝑛
𝐹𝑑(3) (2, 𝑛)
𝐹𝑑(3) (3, 𝑛)
𝐹𝑑(3) (4, 𝑛)
𝐹𝑑(3) (5, 𝑛)
𝐹𝑑(3) (6, 𝑛)
𝐹𝑑(3) (7, 𝑛)
𝐹𝑑(3) (8, 𝑛)
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
2
1
1
1
1
1
1
3
3
3
1
1
1
1
1
4
5
3
3
1
1
1
1
5
8
4
4
3
1
1
1
6
13
6
3
4
3
1
1
7
21
7
4
4
4
3
1
8
34
10
7
3
4
4
3
9
55
13
7
4
4
4
4
10
89
17
7
7
3
4
4
11
144
23
11
8
4
4
4
12
233
30
14
7
7
3
4
13
377
40
14
7
8
4
4
14
610
53
18
11
8
7
3
11
89
7
3
1
1
0
0
12
144
9
3
0
2
0
0
13
233
12
2
1
1
1
0
14
377
16
4
3
0
2
0
Table 4: The 𝑛th distance Fibonacci numbers 𝐹𝑑(4) (π‘˜, 𝑛).
𝑛
𝐹𝑑(4) (2, 𝑛)
𝐹𝑑(4) (3, 𝑛)
𝐹𝑑(4) (4, 𝑛)
𝐹𝑑(4) (5, 𝑛)
𝐹𝑑(4) (6, 𝑛)
𝐹𝑑(4) (7, 𝑛)
𝐹𝑑(4) (8, 𝑛)
0
0
0
0
0
0
0
0
1
1
1
1
1
1
1
1
2
1
0
0
0
0
0
0
3
2
1
0
0
0
0
0
4
3
1
1
0
0
0
0
5
5
1
1
1
0
0
0
6
8
2
0
1
1
0
0
Conflict of Interests
The authors declare that there is no conflict of interests
regarding the publication of this paper.
Acknowledgment
The authors would like to thank the referee for helpful valuable suggestions which resulted in improvements to this paper.
References
[1] U. Bednarz, A. WΕ‚och, and M. WoΕ‚owiec-MusiaΕ‚, “Distance Fibonacci numbers, their interpretations and matrix generators,”
Commentationes Mathematicae, vol. 53, no. 1, pp. 35–46, 2013.
[2] E. KilicΜ§, “The generalized order-π‘˜ Fibonacci-Pell sequence by
matrix methods,” Journal of Computational and Applied Mathematics, vol. 209, no. 2, pp. 133–145, 2007.
[3] I. WΕ‚och and A. WΕ‚och, “Generalized sequences and kindependent sets in graphs,” Discrete Applied Mathematics, vol.
158, no. 17, pp. 1966–1970, 2010.
[4] J. Ercolano, “Matrix generators of Pell sequences,” The Fibonacci
Quarterly, vol. 17, no. 1, pp. 71–77, 1979.
[5] E. Kilic and D. Tasci, “On the generalized Fibonacci and Pell
sequences by Hessenberg matrices,” Ars Combinatoria, vol. 94,
pp. 161–174, 2010.
[6] A. WΕ‚och and M. WoΕ‚owiec-MusiaΕ‚, “Generalized Pell numbers
and some relations with Fibonacci numbers,” Ars Combinatoria,
vol. 109, pp. 391–403, 2013.
[7] D. Jakubczyk, “The exact results in the one-dimensional attractive Hubbard model,” Acta Physica Polonica B, vol. 45, no. 8, pp.
1759–1770, 2014.
[8] M. Startek, A. WΕ‚och, and I. WΕ‚och, “Fibonacci numbers and
Lucas numbers in graphs,” Discrete Applied Mathematics, vol.
157, no. 4, pp. 864–868, 2009.
7
13
2
1
0
1
1
0
8
21
3
2
0
0
1
1
9
34
4
1
1
0
0
1
10
55
5
1
2
0
0
0
Advances in
Operations Research
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Advances in
Decision Sciences
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Journal of
Applied Mathematics
Algebra
Hindawi Publishing Corporation
http://www.hindawi.com
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Journal of
Probability and Statistics
Volume 2014
The Scientific
World Journal
Hindawi Publishing Corporation
http://www.hindawi.com
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
International Journal of
Differential Equations
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Volume 2014
Submit your manuscripts at
http://www.hindawi.com
International Journal of
Advances in
Combinatorics
Hindawi Publishing Corporation
http://www.hindawi.com
Mathematical Physics
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Journal of
Complex Analysis
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
International
Journal of
Mathematics and
Mathematical
Sciences
Mathematical Problems
in Engineering
Journal of
Mathematics
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Discrete Mathematics
Journal of
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Discrete Dynamics in
Nature and Society
Journal of
Function Spaces
Hindawi Publishing Corporation
http://www.hindawi.com
Abstract and
Applied Analysis
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
International Journal of
Journal of
Stochastic Analysis
Optimization
Hindawi Publishing Corporation
http://www.hindawi.com
Hindawi Publishing Corporation
http://www.hindawi.com
Volume 2014
Volume 2014
Download