Y

ITM Web of Conferences 67, 01035 (2024) https://doi.org/10.
1051/itmconf/20246701035
ICMSA 2024
Some Properties of Hyperbolic k-Narayana Quaternions
Tanupat Petpanwong1,∗ and Narawadee Phudolsitthiphat1,∗∗

1
Department of Mathematics, Faculty of Science, Chiang Mai University, Chiang Mai 50200, Thailand
Abstract. In this paper, we introduced the hyperbolic k-Narayana quaternions.
Several properties of these quaternions are investigated, including the Binet for-
mulas, generating functions, and summation formulas. Moreover, some iden-
tities such as Catalan, Cassini, and d’Ocagne are obtained. Our results extend
and generalize well-known theorems.
1 Introduction
Researchers have been studying number sequences for a considerable amount of time. Specif-
ically, the Fibonacci numbers are regarded as the most significant and noteworthy among
these numbers. Over the years, extensive research has been conducted on Fibonacci num-
bers, leading to numerous generalizations. Examples of these generalizations include the
k-Fibonacci, (p, q)-Fibonacci, and Jacobsthal numbers.
There exists a sequence of numbers that defines a recursive relationship similar to the
Fibonacci numbers, known as the Narayana numbers. In the 14th century, Tadepalli Venkata
Narayana introduced them as follows:
Nn = Nn−1 + Nn−3 , n≥3 (1)
with 0,1,1,1 initial condition.
That is, the Narayana sequence is {0, 1, 1, 1, 2, 3, 4, 6, 9, 13, 19, 28, 41, 60, . . . }.
Narayana numbers provide a powerful tool for counting and analyzing combinatorial
structures, and their applications extend to diverse areas within mathematics and computer
science (see [1, 3, 15]).
To generalize the definition of the Narayana numbers, the k-Narayana numbers were in-
troduced in 2015 by Ramírez and Sirvent [14].
Definition 1.1. For a non-zero integer number k, the k-Narayana sequence is given by
Nk,n = kNk,n−1 + Nk,n−3 , n≥3 (2)
with 0, 1, k, k2 initial condition.
The equation x3 +kx2 −1 = 0 corresponds to the characteristic equation of the recurrence rela-
tion (2), with roots denoted as φk , ϖk , ςk . The following equation represents Binet’s formula
for the k-Narayana numbers [14]:
φn+1
k ϖn+1
k ςkn+1
Nk,n = + + , n≥0 (3)
(φk − ϖk )(φk − ςk ) (ϖk − φk )(ϖk − ςk ) (ςk − φk )(ςk − ϖk )
∗ e-mail: tanupat_pet@cmu.ac.th
∗∗ e-mail: narawadee.nanan@cmu.ac.th
© The Authors, published by EDP Sciences. This is an open access article distributed under the terms of the Creative Commons
Attribution License 4.0 (https://creativecommons.org/licenses/by/4.0/).
ITM Web of Conferences 67, 01035 (2024) https://doi.org/10.1051/itmconf/20246701035
ICMSA 2024
Özkan [12] has illustrated the summation formulas for k-Narayana numbers as follows:
n
∑ Nk,1 + Nk,n + Nk,n+1 + Nk,n+2
Nk,i = − Nk,n+1 . (4)
i=0
k
Mathematician Sir William Rowan Hamilton introduced quaternions [8] for the first time
in 1843. Hamilton developed quaternions as a way to extend complex numbers to higher di-
mensions, particularly to three-dimensional space. He was inspired by the complex numbers,
which can be thought of as points on a plane, and sought a similar system for representing
points in three-dimensional space. This led him to the invention of quaternions, which have
applications in various fields, including computer graphics, robotics, and signal processing
(see [2, 4]). Quaternions with specific integer components have been the subject of numer-
ous investigations. The Fibonacci and Lucas quaternions were first introduced and some
properties were provided by Horadam [9] in 1963. Later, Çimen and İpek [5] studied Pell
quaternions and Pell-Lucas quaternions and provided insights into their properties, applica-
tions, and implications in various mathematical fields.
In 1902, Macfarlane [11] introduced the hyperbolic quaternions which, unlike real quater-
nions, do not exhibit commutativity, and conducted a study on their properties. Hyperbolic
quaternions find applications in various fields, including physics, computer graphics, and
geometric algebra (see [6, 16]). A set of hyperbolic quaternions are represented as
H = {h = o + p j1 + x j2 + y j3 : o, p, x, y ∈ R}
where j1 , j2 and j3 are hyperbolic quaternion units satisfying the conditions
j12 = j22 = j32 = j1 j2 j3 = 1, j1 j2 = j3 = − j2 j1 , j2 j3 = j1 = − j3 j2 , j3 j1 = j2 = − j1 j3 .
Let h1 = o1 + p1 j1 + x1 j2 + y1 j3 and h2 = o2 + p2 j1 + x2 j2 + y2 j3 denote two hyperbolic
quaternions. Equality, addition, subtraction, and scalar multiplication and multiplication can
be defined as follows:
h1 = h2 only if o1 = o2 , p1 = p2 , x1 = x2 , y1 = y2 ;
h1 + h2 = (o1 + o2 ) + (p1 + p2 ) j1 + (x1 + x2 ) j2 + (y1 + y2 ) j3 ;
h1 − h2 = (o1 − o2 ) + (p1 − p2 ) j1 + (x1 − x2 ) j2 + (y1 − y2 ) j3 ;
λh1 = λo1 + λp1 j1 + λx1 j2 + λy1 j3 , λ ∈ R.
h1 h2 = (o1 o2 + p1 p2 + x1 x2 + y1 y2 ) + (o1 p2 + p1 o2 + x1 y2 − y1 x2 ) j1
+ (o1 x2 − p1 y2 + x1 o2 + y1 p2 ) j2 + (o1 y2 + p1 x2 − x1 p2 + y1 o2 ) j3 .
Then, the set H of all hyperbolic quaternions is a vector space over a field R. Moreover, the
conjugate of h = o + p j1 + x j2 + y j3 is established by
h̄ = o − p j1 − x j2 − y j3 .
Many researchers have defined and studied various types of hyperbolic quaternions, in-
cluding hyperbolic k-Fibonacci, hyperbolic k-Fibonacci-Lucas, hyperbolic k-Jacobsthal, and
hyperbolic k-Jacobsthal-Lucas quaternions (see [7, 10, 13]).
Several authors worked on the Catalan, Cassini, and d’Ocagne identities because they are
useful in various mathematical contexts, including combinatorics, probability theory, matrix
theory, coding theory, and linear algebra.
The objective of this study is to present the hyperbolic k-Narayana quaternions, including
their generating function, exponential generating function, and Binet’s formula. Addition-
ally, we derive the Catalan, Cassini, and d’Ocagne identities for the hyperbolic k-Narayana
quaternions using Binet’s formula.
2
ICMSA 2024
2 Main Results
In this part, we will initially provide the definition of the hyperbolic k-Narayana quaternions.
Subsequently, we will explore various properties associated with these quaternions.
Definition 2.1. Assume n ≥ 0. The nth k−Narayana hyperbolic quaternions, HNk,n , are
defined by
HNk,n = Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3 (5)
where j1 , j2 , j3 satisfy equalities

j12 = j22 = j32 = j1 j2 j3 = 1, j1 j2 = j3 = − j2 j1 , j2 j3 = j1 = − j3 j2 , j3 j1 = j2 = − j1 j3 .
The first few terms of hyperbolic k-Narayana quaternions are given by
HNk,0 = j1 + k j2 + k2 j3 ,
HNk,1 = 1 + k j1 + k2 j2 + (k3 + 1) j3 ,
HNk,2 = k + k2 j1 + (k3 + 1) j2 + (k4 + 2k) j3 ,
HNk,3 = k2 + (k3 + 1) j1 + (k4 + 2k) j2 + (k5 + 3k2 ) j3 ,
HNk,4 = (k3 + 1) + (k4 + 2k) j1 + (k5 + 3k2 ) j2 + (k6 + 4k3 + 1) j3 ,
..
.
Next, we present the recurrence relations of k-Narayana hyperbolic quaternions.

Lemma 2.2. Assume n ≥ 3, then
HNk,n = kHNk,n−1 + HNk,n−3 . (6)
Proof. Using (5) and (2), we obtain
HNk,n = Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3

= (kNk,n−1 + Nk,n−3 ) + (kNk,n + Nk,n−2 ) j1 + (kNk,n+1 + Nk,n−1 ) j2
+ (kNk,n+2 + Nk,n ) j3
= (kNk,n−1 + kNk,n j1 + kNk,n+1 j2 + kNk,n+2 j3 )
+ (Nk,n−3 + Nk,n−2 j1 + Nk,n−1 j2 + Nk,n j3 )
= k(Nk,n−1 + Nk,n j1 + Nk,n+1 j2 + Nk,n+2 j3 )
+ (Nk,n−3 + Nk,n−2 j1 + Nk,n−1 j2 + Nk,n j3 )
= kHNk,n−1 + HNk,n−3 .
□
The next theorem shows the relationship between hyperbolic quaternions and their conju-
gates.
Lemma 2.3. Assume n ≥ 0, the following equalities are valid:
HNk,n + HNk,n = 2Nk,n , (7)

( )
HNk,n − HNk,n = 2 Nk,n+1 + Nk,n+2 + Nk,n+3 , (8)
HNk,n HNk,n = N2k,n − N2k,n+1 − N2k,n+2 − N2k,n+3 . (9)
3
ICMSA 2024
Proof. It is evident that equations (7)-(8) hold. Next, we will demonstrate the validity of
equation (9). By (5), we have
HNk,n HNk,n
= (Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3 )(Nk,n − Nk,n+1 j1 − Nk,n+2 j2 − Nk,n+3 j3 )
= Nk,n Nk,n − Nk,n Nk,n+1 j1 − Nk,n Nk,n+2 j2 − Nk,n Nk,n+3 j3
+ Nk,n Nk,n+1 j1 − Nk,n+1 Nk,n+1 j1 j1 − Nk,n+1 Nk,n+2 j1 j2 − Nk,n+1 Nk,n+3 j1 j3
= N2k,n − N2k,n+1 − N2k,n+2 − N2k,n+3 .
□
Theorem 2.4. (Binet formula for k-Narayana hyperbolic quaternions) Let φk , ϖk , ςk be the
roots of x3 + kx2 − 1 = 0. Assume n ≥ 0, then
φn+1
k φ̂k ϖn+1
k ϖ̂k ςkn+1 ς̂k
HNk,n = + + , (10)
where φ̂k = 1+φk j1 +φ2k j2 +φ3k j3 , ϖ̂k = 1+ϖk j1 +ϖ2k j2 +ϖ3k j3 , and ς̂k = 1+ςk j1 +ςk2 j2 +ςk3 j3 .
Proof. By using (5) and (3) , we have
HNk,n = Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3

φn+1
k ϖn+1
k ςkn+1
= + +
 
 φn+2
k ϖn+2
k ςkn+2 
+  + +  j1
 
 φn+3
k ϖn+3
k ςkn+3 
+  + +  j2
 
 φn+4
k ϖn+4
k ςkn+4 
+  + +  j3
 
 φn+1
k φn+2
k j1 φn+3
k j2 φn+4
k j3 
=  + + + 
(φk − ϖk )(φk − ςk ) (φk − ϖk )(φk − ςk ) (φk − ϖk )(φk − ςk ) (φk − ϖk )(φk − ςk )
 
 ϖn+1
k ϖn+2
k j1 ϖn+3
k j2 ϖn+4
k j3 
+  + + + 
(ϖk − φk )(ϖk − ςk ) (ϖk − φk )(ϖk − ςk ) (ϖk − φk )(ϖk − ςk ) (ϖk − φk )(ϖk − ςk )
 
 ςkn+1 ςkn+2 j1 ςkn+3 j2 ςkn+4 j3 
+  + + + 
(ςk − φk )(ςk − ϖk ) (ςk − φk )(ςk − ϖk ) (ςk − φk )(ςk − ϖk ) (ςk − φk )(ςk − ϖk )
φn+1
k
( ) ϖn+1
k
( )
= 1 + φk j1 + φ2k j2 + φ3k j3 + 1 + ϖk j1 + ϖ2k j2 + ϖ3k j3
(φk − ϖk )(φk − ςk ) (ϖk − φk )(ϖk − ςk )
ςkn+1 ( )
+ 1 + ςk j1 + ςk2 j2 + ςk3 j3
(ςk − φk )(ςk − ϖk )
φn+1
k φ̂k ϖn+1
k ϖ̂k ςkn+1 ς̂k
= + + .
□
4
ICMSA 2024
Theorem 2.5. The summation for the k−Narayana hyperbolic quaternions is

n
∑ HNk,n + HNk,n+1 + HNk,n+2 − kHNk,n+1 + 1 + j1 + j2 + j3
HNk,m = − j2 −(k+1) j3 . (11)
m=0
k
Proof. By using (5) and (4), we have

n
∑
HNk,m = HNk,0 + HNk,1 + HNk,2 + ... + HNk,n
m=0
= (Nk,0 + Nk,1 j1 + Nk,2 j2 + Nk,3 j3 ) + (Nk,1 + Nk,2 j1 + Nk,3 j2 + Nk,4 j3 )
+ ... + (Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3 )
= (Nk,0 + Nk,1 + Nk,2 + ... + Nk,n ) + (Nk,1 + Nk,2 + Nk,3 + ... + Nk,n+1 ) j1
+ (Nk,2 + Nk,3 + Nk,4 + ... + Nk,n+2 ) j2 + (Nk,3 + Nk,4 + Nk,5 + ... + Nk,n+3 ) j3
n n+1
 n+2   n+3 
∑ ∑ ∑  ∑ 
= Nk,m + j1 Nk,m + j2  Nk,m − 1 + j3  Nk,m − k − 1
m=0 m=0 m=0 m=0
( )
Nk,1 + Nk,n + Nk,n+1 + Nk,n+2 − kNk,n+1
=
k
( )
Nk,1 + Nk,n+1 + Nk,n+2 + Nk,n+3 − kNk,n+2
+ j1
k
( )
+ j2 − j2
k
( )
+ j3 − (k + 1) j3
k
( ) ( )
1 + j1 + j2 + j3 Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3
= +
k k
( )
Nk,n+1 + Nk,n+2 j1 + Nk,n+3 j2 + Nk,n+4 j3
+
k
( )
+
k
( )
−k − j2 − (k + 1) j3
k
HNk,n + HNk,n+1 + HNk,n+2 − kHNk,n+1 + 1 + j1 + j2 + j3
= − j2 − (k + 1) j3 .
k
□
Theorem 2.6. The generating function for k−Narayana hyperbolic quaternions is

∞
∑ HNk,0 + x(HNk,1 − kHNk,0 ) + x2 (HNk,2 − kHNk,1 )
HNk,n xn = . (12)
n=0
1 − kx − x3
Proof. Suppose that the generating function of the k−Narayana hyperbolic quaternions HNk,n
∑
has the form f (x) = ∞ n
n=0 HNk,n x . Then
f (x) = HNk,0 + HNk,1 x + HNk,2 x2 + HNk,3 x3 + ... + HNk,n xn + ...
5
ICMSA 2024
Multiply f (x) on both side by kx and then x3 we have

kx f (x) = kHNk,0 x + kHNk,1 x2 + kHNk,2 x3 + ... + kHNk,n−1 xn + kHNk,n xn+1 + ...
x3 f (x) = HNk,0 x3 + HNk,1 x4 + HNk,2 x5 + ... + HNk,n−1 xn+2 + HNk,n xn+3 + ...
By Lemma 2.2,
(1 − kx − x3 ) f (x) = (HNk,0 + HNk,1 x + HNk,2 x2 + HNk,3 x3 + ... + HNk,n xn + ...)
− (kHNk,0 x + kHNk,1 x2 + kHNk,2 x3 + ... + kHNk,n−1 xn + kHNk,n xn+1 + ...)
− (HNk,0 x3 + HNk,1 x4 + HNk,2 x5 + ... + HNk,n−1 xn+2 + HNk,n xn+3 + ...)
= HNk,0 + x(HNk,1 − kHNk,0 ) + x2 (HNk,2 − kHNk,1 ) + x3 (HNk,3 − (kHNk,2 + HNk,0 ))
+ x4 (HNk,4 − (kHNk,3 + HNk,1 )) + ... + xn (HNk,n − (kHNk,n−1 + HNk,n−3 )) + ...
= HNk,0 + x(HNk,1 − kHNk,0 ) + x2 (HNk,2 − kHNk,1 ) + x3 (HNk,3 − HNk,3 )
+ x4 (HNk,4 − HNk,4 ) + ... + xn (HNk,n − HNk,n )
= HNk,0 + x(HNk,1 − kHNk,0 ) + x2 (HNk,2 − kHNk,1 ).
Therefore,
HNk,0 + x(HNk,1 − kHNk,0 ) + x2 (HNk,2 − kHNk,1 )
f (x) = .
(1 − kx − x3 )
□
Theorem 2.7. For the k−Narayana hyperbolic quaternions, the exponential generating func-
tion is
∑∞
xn φ̂k φk eφk x ϖ̂k ϖk eϖk x ς̂k ςk eςk x
HNk,n = + + . (13)
n=0
n! (φk − ϖk )(φk − ςk ) (ϖk − φk )(ϖk − ςk ) (ςk − φk )(ςk − ϖk )
Proof. By (10), we have

∞ ∞   n
∑ xn ∑  φn+1
k φ̂k ϖn+1
k ϖ̂k ςkn+1 ς̂k  x
HNk,n =  + + 
n=0
n! n=0
(φk − ϖ k )(φ k − ς k ) (ϖ k − φ k )(ϖk − ς k ) (ς k − φ k )(ςk − ϖ k ) n!
∞ ∞ ∞
∑ φn+1
k φ̂k x
n ∑ ϖn+1
k ϖ̂k x
n ∑ ςkn+1 ς̂k xn
= + +
n=0
(φk − ϖk )(φk − ςk )n! n=0 (ϖk − φk )(ϖk − ςk )n! n=0 (ςk − φk )(ςk − ϖk )n!
∞
∑ (φk x)n ∑ (ϖk x)n∞
φk φ̂k ϖk ϖ̂k
= +
(φk − ϖk )(φk − ςk ) n=0 n! (ϖk − φk )(ϖk − ςk ) n=0 n!
∞
∑ (ςk x)n
ςk ς̂k
+ .
(ςk − φk )(ςk − ϖk ) n=0 n!
∑∞ xn
By the exponential generating function n=0 n! = e x , we have
∞
∑ xn φ̂k φk eφk x ϖ̂k ϖk eϖk x ς̂k ςk eςk x
HNk,n = + + .
n=0
n! (φk − ϖk )(φk − ςk ) (ϖk − φk )(ϖk − ςk ) (ςk − φk )(ςk − ϖk )
□
6
ICMSA 2024
Theorem 2.8. (Catalan’s identity) Given two integers, m and n, such that m ≥ n ≥ 0. Then,
φ̂k ς̂k (ςk − φk )2(m−1) (φk ςk )n−m+1 φ̂k ϖ̂k (ϖk − φk )2(m−1) (φk ϖk )n−m+1
HNk,n−m HNk,n+m − HN2k,n = +
(φk − ϖk )(ϖk − ςk ) (ςk − φk )(ϖk − ςk )
ϖ̂k ς̂k (ςk − ϖk )2(m−1) (ϖk ςk )n−m+1
+ .
(φk − ϖk )(ςk − φk )
Proof. By using (10), we have
HNk,n−m HNk,n+m − HN2k,n
 
 φn−m+1
k φ̂k ϖn−m+1
k ϖ̂k ςkn−m+1 ς̂k 
=  + + 
 
 φn+m+1
k φ̂k ϖn+m+1
k ϖ̂k ςkn+m+1 ς̂k 
 + + 
 2
 φn+1
k φ̂k ϖn+1
k ϖ̂k ςkn+1 ς̂k 
−  + + 
φ2n+2
k φ̂2k φn−m+1
k ϖn+m+1
k φ̂k ϖ̂k
= +
(φk − ϖk )(φk − ςk )(φk − ϖk )(φk − ςk ) (φk − ϖk )(φk − ςk )(ϖk − φk )(ϖk − ςk )
φn−m+1
k ςkn+m+1 φ̂k ς̂k ϖn−m+1
k φn+m+1
k ϖ̂k φ̂k
+ +
(φk − ϖk )(φk − ςk )(ςk − φk )(ςk − ϖk ) (ϖk − φk )(ϖk − ςk )(φk − ϖk )(φk − ςk )
ϖ2n+2
k ϖ̂2k ϖn−m+1
k ςkn+m+1 ϖ̂k ς̂k
+ +
(ϖk − φk )(ϖk − ςk )(ϖk − φk )(ϖk − ςk ) (ϖk − φk )(ϖk − ςk )(ςk − φk )(ςk − ϖk )
ςkn−m+1 φn+m+1 k ς̂k φ̂k ςkn−m+1 ϖn+m+1k ς̂k ϖ̂k
+ +
((ςk − φk )(ςk − ϖk )(φk − ϖk )(φk − ςk ) (ςk − φk )(ςk − ϖk )(ϖk − φk )(ϖk − ςk )
ςk2n+2 ς̂k2 φ2n+2
k φ̂2k
+ −
(ςk − φk )(ςk − ϖk )(ςk − φk )(ςk − ϖk ) (φk − ϖk )(φk − ςk )(φk − ϖk )(φk − ςk )
φn+1 n+1
k ϖk φ̂k ϖ̂k φn+1 n+1
k ςk φ̂k ς̂k
− −
(φk − ϖk )(φk − ςk )(ϖk − φk )(ϖk − ςk ) (φk − ϖk )(φk − ςk )(ςk − φk )(ςk − ϖk )
ϖn+1 n+1
k φk ϖ̂k φ̂k ϖ2n+2
k ϖ̂2k
− −
(ϖk − φk )(ϖk − ςk )(φk − ϖk )(φk − ςk ) (ϖk − φk )(ϖk − ςk )(ϖk − φk )(ϖk − ςk )
ϖn+1 n+1
k ςk ϖ̂k ς̂k ςkn+1 φn+1
k ς̂k φ̂k
− −
(ϖk − φk )(ϖk − ςk )(ςk − φk )(ςk − ϖk ) ((ςk − φk )(ςk − ϖk )(φk − ϖk )(φk − ςk )
ςkn+1 ϖn+1 k ς̂k ϖ̂k ςk2n+2 ς̂k2
− −
(ςk − φk )(ςk − ϖk )(ϖk − φk )(ϖk − ςk ) (ςk − φk )(ςk − ϖk )(ςk − φk )(ςk − ϖk )
( 2m
) ( )
(ϖk −φk )2m
φnk ςkn φ̂k ς̂k (ς k −φk )
(φk ςk )m−1
φ n n
ϖ φ̂ ϖ̂
k k k k (φk ϖk )m−1
= +
(φk − ϖk )(φk − ςk )(ςk − φk )(ςk − ϖk ) (φk − ϖk )(φk − ςk )(ϖk − φk )(ϖk − ςk )
( −ϖ )2m )
ϖnk ςkn ϖ̂k ς̂k (ς k k
(ςk ϖk )m−1
+
(ϖk − φk )(ϖk − ςk )(ςk − φk )(ςk − ϖk )
7
ICMSA 2024
φ̂k ς̂k (ςk − φk )2(m−1) (φk ςk )n−m+1 φ̂k ϖ̂k (ϖk − φk )2(m−1) (φk ϖk )n−m+1
= +
(ϖk − φk )(ςk − ϖk ) (ςk − φk )(ϖk − ςk )
ϖ̂k ς̂k (ςk − ϖk )2(m−1) (ϖk ςk )n−m+1
+ .
(φk − ϖk )(ςk − φk )
□
If we set m = 1 in Catalan’s identity, we have Cassini’s identity:
Theorem 2.9. Assume n ≥ 1, the following equality is valid:
φnk ϖnk φ̂k ϖ̂k φnk ςkn φ̂k ς̂k ϖnk ςkn ϖ̂k ς̂k
HNk,n−1 HNk,n+1 − HN2k,n = + + .
(ςk − φk )(ϖk − ςk ) (φk − ϖk )(ϖk − ςk ) (φk − ϖk )(ςk − φk )
The following theorem, we present d’Ocagne’s identity:
Theorem 2.10. Given two integers, m and n, such that m ≥ n ≥ 0. Then,
(ςkn+1 φn+1 m−n
k )ς̂k φ̂k (ςk − φm−n
k ) (φn+1 ϖn+1 m−n
k )φ̂k ϖ̂k (φk − ϖm−n
k )
HNk,m HNk,n+1 − HNk,m+1 HNk,n = + k
(ςk − φk )(ςk − ϖk )(φk − ϖk ) (φk − ϖk )(φk − ςk )(ϖk − ςk )
(ϖn+1 ςn+1 )ϖ̂k ς̂k (ϖm−n − ςkm−n )
+ k k k
.
(ϖk − φk )(ϖk − ςk )(ςk − φk )
Proof. By (10), we obtain
HNk,m HNk,n+1 − HNk,m+1 HNk,n
 
 φm+1
k φ̂k ϖm+1
k ϖ̂k ςkm+1 ς̂k 
=  + + 
 
 φn+2
k φ̂k ϖn+2
 + + 
 
 φm+2
k φ̂k ϖm+2
k ϖ̂k ςkm+2 ς̂k 
−  + + 
 
 φn+1
k φ̂k ϖn+1
 + + 
φm+1 n+2 m+2 n+1
k ϖk φ̂k ϖ̂k − φk ϖk φ̂k ϖ̂k φm+1 n+2 m+2 n+1
k ςk φ̂k ς̂k − φk ςk φ̂k ς̂k
= +
(φk − ϖk )(φk − ςk )(ϖk − φk )(ϖk − ςk ) (φk − ϖk )(φk − ςk )(ςk − φk )(ςk − ϖk )
ϖm+1 n+2 m+2 n+1
k φk ϖ̂k φ̂k − ϖk φk ϖ̂k φ̂k ϖm+1 n+2 m+2 n+1
k ςk φ̂k ς̂k − ϖk ςk ϖ̂k ς̂k
+ +
(φk − ϖk )(φk − ςk )(ϖk − φk )(ϖk − ςk ) (ϖk − φk )(ϖk − ςk )(ςk − φk )(ςk − ϖk )
ςkm+1 φn+2
k ς̂k φ̂k − ςk
m+2 n+1
φk ς̂k φ̂k ςkm+1 ϖn+2
k ς̂k ϖ̂k − ςk
m+2 n+1
ϖk ς̂k ϖ̂k
+ +
(φk − ϖk )(φk − ςk )(ςk − φk )(ςk − ϖk ) (ϖk − φk )(ϖk − ςk )(ςk − φk )(ςk − ϖk )
(φ2k ςk − φk ςk2 )ς̂k φ̂k (ςkm φnk − φm n
k ςk ) (ϖ2k φk − ϖk φ2k )φ̂k ϖ̂k (φm n m n
k ϖk − ϖk φk )
= +
(ςk − φk )(ςk − ϖk )(φk − ϖk )(φk − ςk ) (φk − ϖk )(φk − ςk )(ϖk − φk )(ϖk − ςk )
(ςk2 ϖk − ςk ϖ2k )ϖ̂k ς̂k (ϖm n m n
k ςk − ςk ϖk )
+
(ϖk − φk )(ϖk − ςk )(ςk − φk )(ςk − ϖk )
(ς φ )ς̂k φ̂k (ςkm−n − φm−n
n+1 n+1
k ) (φn+1 ϖn+1 m−n
k )φ̂k ϖ̂k (φk − ϖm−n
k )
= k k + k
(ςk − φk )(ςk − ϖk )(φk − ϖk ) (φk − ϖk )(φk − ςk )(ϖk − ςk )
(ϖn+1 ςn+1 )ϖ̂k ς̂k (ϖm−n − ςkm−n )
+ k k k
.
(ϖk − φk )(ϖk − ςk )(ςk − φk )
□
8
ICMSA 2024
Acknowledgment. The authors would like to thank referees to complete this paper with
valuable suggestions. This research was supported by Chiang Mai University.
References
[1] J. P. Allouche, J. Johnson, Narayanas cows and delayed morphisms, In Articles of 3rd
Computer Music Conference JIM96, France, (1996).
[2] N. Le Bihan, S. J. Sangwine, Quaternion principal component analysis of color images.
InProceedings 2003 International Conference on Image Processing (Cat. No. 03CH37429)
(2003) Sep 14 (Vol. 1, pp. I-809). IEEE.
[3] W. Y. Chen, A general bijective algorithm for trees. Proceedings of the National Academy
of Sciences 87, no. 24 (1990), 9635-9639.
[4] B. Chen, H. Shu and G. Coatrieux, G. Chen, X. Sun, Coatrieux JL. Color image analysis
by quaternion-type moments, Journal of mathematical imaging and vision, Jan;51, (2015)
124-44.
[5] C. B. Çimen and A. İpek, On pell quaternions and Pell-Lucas quaternions. Advances in
Applied Clifford Algebras, 26, (2016), 39-51.
[6] S. Demir, M. Tanl and N. Candemir, Hyperbolic quaternion formulation of electromag-
netism. Advances in Applied Clifford Algebras, 20, (2010), 547-563.
[7] A. Godase, Hyperbolic k-Fibonacci and k-Lucas quaternions, Math. Student 90(1-2)
(2021), 103-116.
[8] W. R. Hamilton, Elements of quaternions, London, Longmans, Green, & Company, 1866.
[9] A. F. Horadam, Complex Fibonacci numbers and Fibonacci quaternions, Amer. Math.
Monthly 70(3) (1963), 289-291.
[10] İ. A. Kösal, A note on hyperbolic quaternions, Universal Journal of Mathematics and
Applications 1(3) (2018), 155-159. https://doi.org/10.32323/ujma.380645
[11] A. Macfarlane, Hyperbolic quaternions, Proc. Roy. Soc. Edinburgh 23 (1902), 169-180.
https: //doi.org/10.1017/S0370164600010385
[12] E. Özkan, B. Kuloǧlu and J. F. Peters, k-Narayana sequence self-simirality, Flip graph
views of k-Narayana self-similarity, Chaos, Solitons & Fractals, Volume 153, Part 2
(2021). https://doi.org/10.1016/j.chaos.2021.111473
[13] E. Özkan, M. Uysal and A. Godase, Hyperbolic k-Jacobsthal and k-Jacobsthal-Lucas
quaternions, Indian J. Pure Appl. Math. 53 (2022), 956967. https://doi.org/10.1007/
s13226-021-00202-9
[14] J. L. Ramírez and V. F. Sirvent, A note on the k-Narayana sequences, Ann. Math. Inf.,
45 (2015), 91-105.
[15] R. A. Sulanke, Constraint-sensitive Catalan path statistics having the Narayana distri-
bution. Discrete mathematics, 204(1-3), (1999), 397-414.
[16] K. Takahashi, Comparison of high-dimensional neural networks using hypercomplex
numbers in a robot manipulator control. Artificial Life and Robotics. 2021 Aug;26(3):367-
77.
[17] T. Yaǧmur, A note on hyperbolic (p, q)−Fibonacci quaternions, Commun. Fac. Sci.
Univ. Ank. Ser. A1 Math. Stat., Volume 69, Number 1, Pages 880-890 (2020), DOI:
10.31801.

Y

Uploaded by

Y

Uploaded by

ITM Web of Conferences 67, 01035 (2024) https://doi.org/10.

Some Properties of Hyperbolic k-Narayana Quaternions

Tanupat Petpanwong1,∗ and Narawadee Phudolsitthiphat1,∗∗

HNk,n = Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3 (5)

where j1 , j2 , j3 satisfy equalities

Next, we present the recurrence relations of k-Narayana hyperbolic quaternions.

HNk,n = kHNk,n−1 + HNk,n−3 . (6)

Proof. Using (5) and (2), we obtain

HNk,n = Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3

HNk,n + HNk,n = 2Nk,n , (7)

HNk,n = Nk,n + Nk,n+1 j1 + Nk,n+2 j2 + Nk,n+3 j3

Theorem 2.5. The summation for the k−Narayana hyperbolic quaternions is

Proof. By using (5) and (4), we have

Theorem 2.6. The generating function for k−Narayana hyperbolic quaternions is

f (x) = HNk,0 + HNk,1 x + HNk,2 x2 + HNk,3 x3 + ... + HNk,n xn + ...

Multiply f (x) on both side by kx and then x3 we have

Proof. By (10), we have

Proof. By using (10), we have

HNk,n−m HNk,n+m − HN2k,n

You might also like