See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/281108357

q−Bernoulli Matrices and Their Some Properties

Article  in  Gazi University Journal of Science · July 2015

CITATIONS READS

2 168

2 authors:

Naim Tuglu Semra Kus

Gazi University Ahi Evran University,Kırşehir,Turkey
43 PUBLICATIONS   237 CITATIONS    5 PUBLICATIONS   3 CITATIONS   

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Naim Tuglu on 20 August 2015.

The user has requested enhancement of the downloaded file.

 Gazi University Journal of Science
GU J Sci
28(2):269-273 (2015)

𝑞 −Bernoulli Matrices and Their Some Properties

Naim TUGLU1,♠, Semra KUŞ2

1
Gazi University, Department of Mathematics 06500, Ankara-Turkey
2
Ahi Evran University, Mucur Vocational High School 40500, Kırşehir-Turkey

Received: 24/12/204 Accepted:13/01/2015

ABSTRACT

In this study, we define 𝑞 −Bernoulli matrix B (q) and 𝑞 −Bernoulli polynomial matrix B ( x, q) by using
𝑞 −Bernoulli numbers, and polynomials respectively. We obtain some properties of B (q) and B ( x, q) . We
obtain factorizations 𝑞 −Bernoulli polynomial matrix and shifted 𝑞 −Bernoulli matrix using special matrices.
Keywords: 𝑞 −Bernoulli numbers, 𝑞 −Bernoulli matrix, 𝑞 −Vandermonde matrix.

1. INTRODUCTION 𝑛 [𝑛]𝑞 !
( ) =
𝑘 𝑞 [𝑛 − 𝑘]𝑞 ! [𝑘]𝑞 !
Bernoulli numbers are defined by Jacob Bernoulli ([1]).
Nörlund ([2]) and Carlitz ([3]) obtained some properties of 𝑛 𝑛
with ( ) = 1 and ( ) = 0 for 𝑛 < 𝑘 ([6]). Some
Bernoulli numbers and polynomials. Carlitz ([4, 5]) defined 0 𝑞 𝑘 𝑞
𝑞 −Bernoulli numbers and polynomials. Hegazi ([10]) studied properties of 𝑞 −binomial coefficients are
𝑞 −Bernoulli numbers and polynomials.
𝑛 𝑛
( ) =( ) (1.1)
Let 𝑛 be a positive integer and 𝑞 ∈ (0,1) . The quantum 𝑘 𝑞 𝑛−𝑘 𝑞
integer or Gauss number [𝑛]𝑞 is defined by
and
𝑞𝑛 − 1
[𝑛]𝑞 = = 1 + 𝑞 + 𝑞2 + ⋯ + 𝑞𝑛−1 . 𝑛 𝑘 𝑛 𝑛−𝑗
𝑞−1 ( ) ( ) = (𝑗) ( ) . (1.2)
𝑘 𝑞 𝑗 𝑞 𝑞 𝑘−𝑗 𝑞
The 𝑞 −analogue of 𝑛! is defined as follows
The 𝑞 −analogue of (𝑥 − 𝑎)𝑛 denoted (𝑥 − 𝑎)𝑛𝑞 is
1 𝑖𝑓 𝑛 = 0,
[𝑛]𝑞 ! = { 1 𝑖𝑓 𝑛 = 0,
[𝑛]𝑞 [𝑛 − 1]𝑞 ⋯ [1]𝑞 𝑖𝑓 𝑛 = 1,2, … . (𝑥 − 𝑎)𝑛𝑞 = {
(𝑥 − 𝑎)(𝑥 − 𝑞𝑎) ⋯ (𝑥 − 𝑞𝑛−1 𝑎) 𝑖𝑓 𝑛 = 1,2, … .
Gaussian or 𝑞 −binomial coefficients are defined for integers
𝑛 ≥ 𝑘 ≥ 1 as for 𝑥 variable. Using definition of 𝑞 −binomial coefficients it
can be obtained

♠
Corresponding author, e-mail: naimtuglu@gazi.edu.tr
2
270 GU J Sci, 28(2):269-273 (2015)/ Naim TUGLU, Semra KUŞ

𝑛
𝑛 𝑘
𝑞 −Bernoulli polynomials.
(𝑥 + 𝑎)𝑛𝑞 = ∑ ( ) 𝑞 (2) 𝑎𝑘 𝑥 𝑛−𝑘 (1.3)
𝑘 𝑞
𝑘=0 Definition 2. [10] Let 𝑛 be a nonnegative integer and 𝐵𝑛 be
𝑛𝑡ℎ Bernoulli numbers. The 𝑞 −Bernoulli numbers 𝑏𝑛 (𝑞) are
is called Gauss’s binomial formula. defined by
[𝑛]𝑞 !
2. BERNOULLI NUMBERS AND POLYNOMIALS 𝑏𝑛 (𝑞) = 𝐵𝑛 . (2.6)
𝑛!
Firstly, we mention that Bernoulli numbers. Then using these The 𝑞 −Bernoulli polynomials 𝐵𝑛 (𝑥, 𝑞) are defined by
numbers, a matrix can be delivered. This matrix is called
Bernoulli matrix. Extending this matrix some matrices are 𝑛
𝑛
obtained. 𝐵𝑛 (𝑥, 𝑞) = ∑ ( ) 𝑏𝑘 (𝑞)𝑥 𝑛−𝑘 . (2.7)
𝑘 𝑞
𝑘=0
In [7], the Bernoulli numbers are defined initial condition by
𝐵0 = 1 and Theorem 1. [10] For 𝑞 −commuting variables 𝑥 and 𝑦 such
that 𝑥𝑦 = 𝑞𝑥𝑦 we have
𝑛−1
1 𝑛+1
𝐵𝑛 = − ∑( ) 𝐵𝑘 𝑛 = 1,2,3, … . (2.1) 𝑛
𝑛+1 𝑘 𝑛
𝑘=0 𝐵𝑛 (𝑥 + 𝑦, 𝑞) = ∑ ( ) 𝑦 𝑛−𝑘 𝐵𝑘 (𝑥, 𝑞) . (2.8)
𝑘 𝑞
𝑘=0
The exponential generating function of Bernoulli numbers is
Similar considerations apply this theorem, it can easy to check
∞
𝑡 𝑡𝑛 that
= ∑ 𝐵𝑛 . (2.2)
𝑒𝑡 − 1 𝑛! 𝑛
𝑛=0
𝑛
𝐵𝑛 (𝑥 + 𝑦, 𝑞) = ∑ ( ) 𝑥 𝑛−𝑘 𝐵𝑘 (𝑦, 𝑞) . (2.9)
Let n be a nonnegative integer, the Bernoulli polynomials 𝑘 𝑞
𝑘=0
𝐵𝑛 (𝑥) are defined by
3. 𝒒 −BERNOULLI MATRICES
∞
𝑛
𝐵𝑛 (𝑥) = ∑ ( ) 𝐵𝑘 𝑥 𝑛−𝑘 . (2.3) Zhang [8] defined generalized Bernoulli matrix by using
𝑘
𝑛=0 Bernoulli numbers and polynomials. Then the author obtained
factorization and some properties of the Bernoulli matrices.
Zhang defined Bernoulli matrices by using Bernoulli numbers
and polynomials. Also the author obtained factorization and Ernst [9] studied matrix form of 𝑞 −Bernoulli polynomials and
some properties of Bernoulli matrices [8]. obtained recurrence formula using this matrix form. The author
studied relation between 𝑞 −Cauchy-Vandermonde matrix and
Definition 1. [8] Let 𝐵𝑛 be 𝑛𝑡ℎ Bernoulli number and 𝐵𝑛 (𝑥) the 𝑞 − Bernoulli matrix. Then the author obtained
be Bernoulli polynomial, (𝑛 + 1) × (𝑛 + 1) type Bernoulli 𝑞 − analogue of the Bernoulli theorem by using the
matrix B  [𝑏𝑖𝑗 ] and Bernoulli polynomial matrix Jackson-Hahn-Cigler 𝑞 −Bernoulli polynomials.
B ( x)  [𝑏𝑖𝑗 (𝑥)] defined respectively as follows In this section, we define 𝑞 −Bernoulli matrices by using
𝑞 −Bernoulli numbers and 𝑞 −Bernoulli polynomials, Then we
𝑖 obtain inverse of 𝑞 − Bernoulli matrix and some theorems
( ) 𝐵𝑖−𝑗 𝑖𝑓 𝑖 ≥ 𝑗,
𝑏𝑖𝑗 = { 𝑗 (2.4) related to the generalized 𝑞 −Bernoulli matrix.
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒,
Definition 3. Let 𝑏𝑛 (𝑞) be 𝑛𝑡ℎ 𝑞 −Bernoulli number. The
and 𝑞 −Bernoulli matrix B (q)  [𝑏𝑖𝑗 (𝑞)] is defined by

𝑖 𝑖
( ) 𝐵𝑖−𝑗 (𝑥) 𝑖𝑓 𝑖 ≥ 𝑗, ( ) 𝑏𝑖−𝑗 (𝑞) 𝑖𝑓 𝑖 ≥ 𝑗,
𝑏𝑖𝑗 (𝑥) = { 𝑗 (2.5) 𝑗
𝑏𝑖𝑗 (𝑞) = { 𝑞 (3.1)
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒. 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒,
It is know that the constant terms of 𝐵𝑛 (𝑥) Bernoulli
where 0 ≤ 𝑖, 𝑗 ≤ 𝑛.
polynomials are 𝐵𝑛 Bernoulli numbers. Therefore we obtain
Bernoulli B matrix by using the constant term of B ( x) 5 × 5 𝑞 −Bernoulli matrix is
Bernoulli polynomial matrix [8].

Now we give definitions of 𝑞 − Bernoulli numbers and

 GU J Sci, 28(2):269-273 (2015)/ Naim TUGLU, Semra KUŞ 271

1 0 0 0 0 Definition 4. Let 𝐵𝑛 (𝑥, 𝑞) be 𝑛𝑡ℎ 𝑞 −Bernoulli polynomial.

1
−
2
1 0 0 0 The 𝑞 −Bernoulli polynomial matrix B ( x, q)  [𝑏𝑖𝑗 (𝑥, 𝑞)] is
[2]𝑞 [2]𝑞
− 1 0 0 defined as follows
B ( q)  2∙3! 2!
[3]𝑞 ∙[2]𝑞 [3]𝑞
0 2∙3!
−
2!
1 0 𝑖
( ) 𝐵𝑖−𝑗 (𝑥, 𝑞) 𝑖𝑓 𝑖 ≥ 𝑗,
[4]𝑞
0
[4]𝑞 ∙[3]𝑞 [4]𝑞
1) 𝑏𝑖𝑗 (𝑥, 𝑞) = { 𝑗 𝑞 (3.3)
( −
30∙4! 2∙3!
−
2!
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.
Following theorem is a generalization of Theorem 2.4 in [8].
4. 𝒒 −BERNOULLI AND 𝒒 −PASCAL MATRICES
Theorem 2. Let D (q)  [𝑑𝑖𝑗 (𝑞)] be (𝑛 + 1) × (𝑛 + 1) Ernst [9] defined (𝑛 + 1) × (𝑛 + 1) generalized 𝑞 − Pascal
matrix, is defined by matrix P ( x, q)  [𝑝𝑖𝑗 (𝑞)] by
𝑖 [𝑖 − 𝑗]𝑞 !
( ) 𝑖𝑓 𝑖 ≥ 𝑗, 𝑖
𝑑𝑖𝑗 (𝑞) = { 𝑗 𝑞 (𝑖 − 𝑗 + 1)! (3.2) ( ) 𝑥 𝑖−𝑗 𝑖𝑓 𝑖 ≥ 𝑗,
𝑝𝑖𝑗 (𝑞) = { 𝑗 𝑞 (4.1)
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.
Then D (q) is the inverse of 𝑞 −Bernoulli matrix. The inverse of generalized 𝑞− Pascal matrix
′
P 1
( x, q)  𝑖𝑗
[𝑝 (𝑞)] is
Proof. Let B (q) be 𝑞 − Bernoulli matrix and D (q)
defined as in (3.2). 𝑖 𝑖−𝑗
( ) 𝑞( 2 ) (−𝑥)𝑖−𝑗 𝑖𝑓 𝑖 ≥ 𝑗,
′
𝑝𝑖𝑗 (𝑞) = { 𝑗 𝑞 (4.2)
𝑛
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.
B (q)D (q) i j = ∑ 𝑏𝑖𝑘 (𝑞)𝑑𝑘𝑗 (𝑞)
𝑘=0 Now using the Zhang’s methods in [8] we can generalize the
𝑖
[𝑘 − 𝑗]𝑞 ! factorization 𝑞 −Bernoulli matrices.
𝑖 𝑘
= ∑ ( ) 𝑏𝑖−𝑘 (𝑞) ( )
𝑘 𝑞 𝑗 𝑞 (𝑘 − 𝑗 + 1)!
𝑘=𝑗 Theorem 3. Let B ( x, q) be 𝑞 −Bernoulli polynomial matrix
𝑖
𝑖 𝑘 [𝑘 − 𝑗]𝑞 ! and P ( x, q) be generalized 𝑞 −Pascal matrix, then
= ∑( ) ( ) 𝑏 (𝑞)
𝑘 𝑞 𝑗 𝑞 (𝑘 − 𝑗 + 1)! 𝑖−𝑘
𝑘=𝑗
𝑖
B ( x  y, q)  P ( y, q) B ( x, q)  P ( x, q) B ( y, q) (4.3)
𝑖 𝑖−𝑗 [𝑘 − 𝑗]𝑞 !
= ( ) ∑( ) 𝑏 (𝑞)
𝑗 𝑞 𝑘 − 𝑗 𝑞 (𝑘 − 𝑗 + 1)! 𝑖−𝑘 and specially
𝑘=𝑗
𝑖−𝑗
𝑖 𝑖−𝑗 [𝑡]𝑞 ! B ( x, q)  P ( x, q) B (q) . (4.4)
= ( ) ∑( ) 𝑏 (𝑞)
𝑗 𝑞 𝑡 𝑞 (𝑡 + 1)! 𝑖−𝑗−𝑡
𝑡=0
Proof. Let P ( y, q) be generalized 𝑞 − Pascal matrix and
𝑖−𝑗
𝑖 𝑖−𝑗 [𝑡]𝑞 ! [𝑖 − 𝑗 − 𝑡]𝑞 ! B ( x, q) be 𝑞 −Bernoulli polynomial matrix. Then
= ( ) ∑( ) 𝐵
𝑗 𝑞 𝑡 𝑞 (𝑡 + 1)! 𝑖−𝑗−𝑡 (𝑖 − 𝑗 − 𝑡)!
𝑡=0
𝑛
𝑖−𝑗
𝑖 [𝑖 − 𝑗]𝑞 ! 𝑖−𝑗 1 P ( y, q) B ( x, q) i j = ∑ 𝑝𝑖𝑘 (𝑞) 𝑏𝑘𝑗 (𝑥, 𝑞)
=( ) ∑( ) 𝐵
𝑗 𝑞 (𝑖 − 𝑗)! 𝑡 𝑞 𝑡 + 1 𝑖−𝑗−𝑡 𝑘=0
𝑖
𝑡=0
𝑖 𝑘
= ∑ ( ) 𝑦 𝑖−𝑘 ( ) 𝐵𝑘−𝑗 (𝑥, 𝑞)
𝑘 𝑞 𝑗 𝑞
Using the orthogonality relation for Bernoulli numbers 𝑘=𝑗
𝑖
𝑖 𝑖−𝑗
𝑛 = ∑( ) ( ) 𝑦 𝑖−𝑘 𝐵𝑘−𝑗 (𝑥, 𝑞)
𝑛 1 𝑗 𝑞 𝑘−𝑗 𝑞
∑( ) 𝐵 = 𝛿𝑛,0 (3.3) 𝑘=𝑗
𝑘 𝑘 + 1 𝑛−𝑘 𝑖−𝑗
𝑘=0
𝑖 𝑖−𝑗
= ( ) ∑( ) 𝑦 𝑖−𝑗−𝑡 𝐵𝑡 (𝑥, 𝑞).
(see [8]), we obtain 𝑗 𝑞 𝑡 𝑞
𝑡=0

𝑖 [𝑖 − 𝑗]𝑞 !
B (q)D (q) i j =( ) 𝛿
𝑗 𝑞 (𝑖 − 𝑗)! 𝑖−𝑗,0
= 𝛿𝑖,𝑗 .
Using (2.8), we have
272 GU J Sci, 28(2):269-273 (2015)/ Naim TUGLU, Semra KUŞ

𝑖
P ( y, q) B ( x, q) i j = ( ) 𝐵𝑖−𝑗 (𝑥 + 𝑦, 𝑞) =
𝑗 𝑞  B ( x  y, q) i j Corollary 1. Let B ( x, q) be 𝑞 −Bernoulli polynomial matrix
then B 1
( x, q)  [𝑐𝑖𝑗 (𝑞)] is
and similarly it can be provide that
𝑖−𝑗 𝑡
[𝑖]𝑞 ! 𝑞 (2) (−𝑥)𝑡
B ( x  y, q)  P ( x, q) B ( y, q) . 𝑐𝑖𝑗 (𝑞) = {[ 𝑗]𝑞 ! ∑ 𝑖𝑓 𝑖 ≥ 𝑗,
[𝑡]𝑞 ! (𝑖 − 𝑗 − 𝑡 + 1)! (4.5)
𝑡=0

Now, we show that 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒.

Proof. Let B (q) be 𝑞 −Bernoulli matrix and P ( x, q) be

B ( x, q)  P ( x, q) B (q) .
generalized 𝑞 −Pascal matrix. Using factorization of B ( x, q)
𝑛
in (4.4)
P ( x, q) B (q) i j = ∑ 𝑝𝑖𝑘 (𝑞) 𝑏𝑘𝑗 (𝑞)
𝑘=0
𝑖
B 1 ( x, q)  B 1 (q) P 1
( x, q)  D (q) P 1
( x, q)
𝑖 𝑘
= ∑ ( ) 𝑥 𝑖−𝑘 ( ) 𝑏𝑘−𝑗 (𝑞)
𝑘 𝑞 𝑗 𝑞 and inverse of generalized 𝑞 −Pascal matrix (4.2), we obtain
𝑘=𝑗
𝑖
𝑖 𝑖−𝑗 𝑛
= ( ) ∑(
𝑗 𝑞 𝑘−𝑗 𝑞
) 𝑥 𝑖−𝑘 𝑏𝑘−𝑗 (𝑞)
𝑘=𝑗
D (q) P 1
( x, q) 
ij
′
= ∑ 𝑑𝑖𝑘 (𝑞)𝑝𝑘𝑗 (𝑞)
𝑘=0
𝑖 𝑖
[𝑖 − 𝑘]𝑞 !
= ( ) 𝐵𝑖−𝑗 (𝑥, 𝑞) 𝑖 𝑘−𝑗 𝑘
(−𝑥)𝑘−𝑗 𝑞( 2 ) ( )
𝑗 𝑞 = ∑( )
𝑗 𝑞
𝑘 𝑞 (𝑖 − 𝑘 + 1)!
 B ( x, q) i j
𝑘=𝑗
= 𝑖 (𝑘−𝑗)
𝑖 𝑖 − 𝑗 [𝑖 − 𝑘]𝑞 ! 𝑞 2
= ∑( ) ( ) (−𝑥)𝑘−𝑗
𝑗 𝑞 𝑘 − 𝑗 𝑞 (𝑖 − 𝑘 + 1)!
𝑘=𝑗
We give two examples of this theorem for 3 × 3 and 4 × 4 𝑖−𝑗
(𝑡)
𝑞 −Bernoulli polynomial matrix and 𝑞 −Pascal matrix. 𝑖 𝑖 − 𝑗 [𝑖 − 𝑗 − 𝑡]𝑞 ! 𝑞 2 (−𝑥)𝑡
= ( ) ∑( )
𝑗 𝑞 𝑡 𝑞 (𝑖 − 𝑗 − 𝑡 + 1)!
𝑡=0

P ( y, q) B ( x, q) i j
𝑖−𝑗 𝑡
[𝑖]𝑞 ! 𝑞(2) (– 𝑥)𝑡
= ∑
1 0 0 1 0 0 [ 𝑗]𝑞 ! [𝑡]𝑞 ! (𝑖 − 𝑗 − 𝑡 + 1)!
𝑡=0

1
= 𝑐𝑖𝑗 (𝑞)
= 𝑦 1 0 × 𝑥−2 1 0

2 2 [2]𝑞 [2]𝑞
[2]𝑞 𝑥 −
[2]𝑞 5. SHIFTED 𝒒 −BERNOULLI AND 𝒒 −VANDERMONDE
(𝑦 [2]𝑞 𝑦 1) (𝑥 − 2
𝑥+ 12 2
1)
1 0 0 MATRICES

= (𝑥 + 𝑦) − 1 1 0 In [8] Zhang defined shifted Bernoulli matrix, and obtained

2
some relations between shifted Bernoulli matrix and
2 2 [2]𝑞 [2]𝑞 [2]𝑞
(𝑥 + [2]𝑞 𝑥𝑦 + 𝑦 − 2
(𝑥 + 𝑦) +
12
[2]𝑞 (𝑥 + 𝑦) −
2
1) Vandermonde matrix.
= B ( x  y, q) In this section we define 𝑞 − Vandermonde matrix. and
𝑞 − shifted Bernoulli matrix by using 𝑞 − Bernoulli
polynomials and give its relation with 𝑞 −Vandermonde matrix.
P ( x, q) B (q) i j
1 0 0 0 1 0 0 0 Definition 5. Let 𝐵𝑛 (𝑥, 𝑞) be 𝑞 −Bernoulli polynomial. The
1 ~
𝑥 1 0 0 −
2
1 0 0 shifted 𝑞 −Bernoulli matrix B ( y, q)  [𝑏̃𝑖𝑗 (𝑦, 𝑞)] is defined
= × [2]𝑞 [2]𝑞
𝑥2 [2]𝑞 𝑥 1 0 − 1 0 by
2∙3! 2!

𝑏̃𝑖𝑗 (𝑦, 𝑞) = 𝐵𝑖 (𝑦 + 𝑗, 𝑞)
[2]𝑞 [3]𝑞 [3]𝑞
( 𝑥3 [3]𝑞 𝑥 2 [3]𝑞 𝑥 1) (0 2∙3!
−
2!
1) (5.1)

1 0 0 0 where 0 ≤ 𝑖, 𝑗 ≤ 𝑛.
1
𝑥− 2
1 0 0 Definition 6. ([11]) The (𝑛 + 1) × (𝑛 + 1) type
=
𝑥 − 2 [2]𝑞
𝑥+
[2]𝑞
[2]𝑞 𝑥 −
[2]𝑞
1 0 𝑞 −Vandermonde matrix V( y, q)  [𝑣𝑖𝑗 (𝑦, 𝑞)] is defined by
2 12 2

3 [3]𝑞 2 [2]𝑞 [3]𝑞 [2]𝑞 [3]𝑞 [2]𝑞 [3]𝑞 [3]𝑞

(𝑥 − 𝑥 + 𝑥 [3]𝑞 𝑥 2 − 𝑥+ [3]𝑞 𝑥 − 1) 𝑣𝑖𝑗 (𝑦, 𝑞) = (𝑦 + 𝑗)𝑖𝑞 . (5.2)
2 12 2 12 2

= B ( x, q)
 GU J Sci, 28(2):269-273 (2015)/ Naim TUGLU, Semra KUŞ 273

4 × 4 𝑞 −Vandermonde matrix is REFERENCES

1 1 1 1 [1] Bernoulli, J., “Ars conjectandi”, Published Posthumously

𝑦 𝑦+1 𝑦+2 𝑦+3 Basel, Switzerland, 1-33, (1713).
V ( y, q) = 𝑦2 (𝑦 + 1)2𝑞 (𝑦 + 2)2𝑞 (𝑦 + 3)2𝑞 .
[2] Nörlund, N. E., “Vorlesungen uber dierenzenrechnung”,
𝑦3 (𝑦 + 1)3𝑞 (𝑦 + 2)3𝑞 (𝑦 + 3)3𝑞
( ) Chelsea Publishing Company, New York, (1924).

In the following theorem, we obtain factorization shifted [3] Carlitz, L., “Some theorems on Bernoulli Numbers of higher
𝑞 − Bernoulli matrix by using 𝑞 − Bernoulli matrix and order received”, Pacic J. Math., 2: 127-139 (1952).
𝑞 −Vandermonde matrix.
[4] Carlitz, L., “𝑞 −Bernoulli numbers and polynomials”, Duke
Theorem 4. Let V( y, q) be 𝑞 − Vandermonde matrix and Math. J., 15 (4): 987-1000 (1948).
B (q) be 𝑞 −Bernoulli matrix. Then [5] Carlitz, L., “Expansions of 𝑞 −Bernoulli numbers”, Duke
Math. J., 25, 355-364 (1958).
~
B ( y, q)  B (q)V( y, q) . [6] Kac, V., Cheung P., “Quantum Calculus”, Springer, New
York (2002)
Proof.
𝑛
[7] Lalin, M. N., “Bernoulli numbers” Junior Number Theory
Seminar-Universty of Texas at Austin September 6th (2005)
 B (q)V( y, q) i j = ∑ 𝑏𝑖𝑘 (𝑞) 𝑣𝑘𝑗 (𝑦, 𝑞)
𝑘=0 [8] Zhang, Z.,Whang,J., “Bernoulli matrix and its algebraic
𝑖
𝑖 properties” Discrete Appl. Math., 154: 1622-1632 (2006).
= ∑ ( ) 𝑏𝑖−𝑘 (𝑞) (𝑦 + 𝑗)𝑘𝑞
𝑘 𝑞
𝑘=0 [9] Ernst,T., “𝑞 −Pascal and 𝑞 −Bernoulli matrices and umbral
approach”, Department of Mathematics Uppsala Universty
If we use definition of 𝑞 −Bernoulli polynomial, then
D.M. Report 2008:23 (2008).

 B (q)V( y, q) i j = 𝐵𝑖 (𝑦 + 𝑗, 𝑞) [10] Hegazi, A. S. and Mansour, M., “A note on 𝑞 −Bernoulli

numbers and polynomials”, J. Nonlinear Math.
we obtain Phys.,13(1):9-18(2005).

~ [11] Song, S.-Z., Cheon, G.-S., Jun, Y.-B., and Beasley, L.-B.,
B (q)V( y, q)  B ( y, q) . “A 𝑞 − analogue of the generalized factorial numbers”, J.
Korean Math. Soc., 47, 645-657 (2010).
For 𝑞 → 1− , we can obtain Theorem 5.2 in [8].

The factorization of 3 × 3 shifted 𝑞 −Bernoulli matrix is as

follows.

B (q)V( y, q)
1 0 0 1 1 1
1
= ( −2 1 0) ( 𝑦 𝑦+1 𝑦+2 )
[2]𝑞 [2]𝑞
2∙3!
−
2!
1 𝑦2 (𝑦 + 1)2𝑞 (𝑦 + 2)2𝑞
1 1 1
1 1 3
=( 𝑦−2 𝑦+2 𝑦+2 )
2 [2] [2] [2]𝑞 7∙[2]𝑞 3∙[2]𝑞 37∙[2]𝑞
𝑦 − 2 𝑞𝑦 + 2∙3!𝑞 𝑦2 + 2
𝑦+ 2∙3!
− 1 𝑦2 + 2
𝑦+ 2∙3!
−− 4

~
= B ( y, q) .

CONFLICT OF INTEREST

The authors declare that there is no conflict of interests

regarding the publication of this paper.

View publication stats