LINEAR ALGEBRA OF PASCAL MATRICES
LINDSAY YATES
Abstract. The famous Pascal’s triangle appears in many areas of
mathematics, such as number theory, combinatorics and algebra.
Pascal matrices are derived from this triangle of binomial coeffi-
cients, which create simplistic matrices with interesting properties.
We explore properties of these matrices and the inverse of the Pascal
matrix plus the identity matrix times any positive integer. We
further consider a unique matrix called the Stirling matrix, which
can be factorized in terms of the Pascal matrix.
1. Introduction
The ancient arithmetic triangle, today known as Pascal’s triangle, is
an infinite numerical table represented in triangular form. The numbers
displayed in the triangle are called binomial coefficients,
n
k
, which
represent the number of ways of picking k unordered outcomes from n
possibilities. Each entry in the triangle is obtained by adding together
two entries from the row above: the one directly to the left and the
one directly to the right; this pattern can be seen in the image below.
The Pascal’s triangle has been known for over ten centuries. The set
of numbers that form the Pascal’s triangle were known before Blaise
Date: December 5, 2014.
1
LINEAR ALGEBRA OF PASCAL MATRICES 2
Pascal, although he is attributed with being the first one to publish
the information known about the triangle in his treatise, Trait´e du triangle
arithm´etique. The numbers originally arose from Indian studies
of combinatorics and the Greeks interest in figurate numbers. These
numbers were continually discussed by Islamic mathematicians during
the 10th century and in the 11th century by a Persian poet named
Omar Khayyam. They were also seen in China during the 13th century.
The Pascal’s triangle was officially published in Pascal’s treatise
soon after his death in 1665.
This triangle arises in many areas of mathematics such as algebra,
probability, and combinatorics. We were motivated by the Pascal’s
triangle prominence in the field of mathematics and its many applications,
in particular Pascal matrices. We wanted to further our studies
to consider the various properties and unique connections that Pascal
matrices has to other functions and number sequences.
2. Pascal Matrices
The Pascal’s triangle can be transcribed into a matrix containing
the binomial coefficients as its elements. We can form three types of
matrices: symmetric, lower triangular, and upper triangular, for any
integer n > 0.
The symmetric Pascal matrix of order n is defined by Sn = (sij ), where
sij =
i + j − 2
j − 1
for i, j = 1, 2, ...., n (1)
We can define the lower triangular Pascal matrix of order n by Ln =
(lij ), where
lij =
(
i−1
j−1
if i ≥ j
0 otherwise
(2)
The upper triangular Pascal matrix of order n is defined by Un = (uij ),
where
uij =
(
j−1
i−1
if j ≥ i
0 otherwise
(3)
LINEAR ALGEBRA OF PASCAL MATRICES 3
We notice that Un = (Ln)
T
, for any positive integer n.
For example, for n = 5 we have:
S5 =
1 1 1 1 1
1 2 3 4 5
1 3 6 10 15
1 4 10 20 35
1 5 15 35 70
L5 =
1 0 0 0 0
1 1 0 0 0
1 2 1 0 0
1 3 3 1 0
1 4 6 4 1
U5 =
1 1 1 1 1
0 1 2 3 4
0 0 1 3 6
0 0 0 1 4
0 0 0 0 1
These Pascal matrices have some interesting properties, which we present
next.
Theorem 2.1. [1] Let Sn be the symmetric Pascal matrix of order n
defined by (1), Ln be the lower triangular Pascal matrix of order n de-
fined by (2), and Un be the upper triangular Pascal matrix of order n
defined by (3), then Sn = LnUn.
Proof. Let Ln be the lower triangular Pascal matrix of order n defined
by (2) and Un be the upper triangular Pascal matrix of order n defined
by (3). By direct multiplication of matrices Ln and Un we obtain the
ij-th element of the product LnUn:
Xn
k=1
likukj =
Xn
k=1
liklkj , since Un = (Ln)
T
.
Then, Xn
k=1
likljk =
Xn
k=1
i − 1
k − 1
j − 1
k − 1
=
X
j
k=1
i − 1
k − 1
j − 1
k − 1
=
=
X
j
k=1
i − 1
k − 1
j − 1
j − k
, since lik = 0 for k>j.
The Vandermonde identity says that:
Xn
t=0
m
t
n
n − t
=
m + n
n
, for any m,n,t ∈ N (4)
Let m = i − 1, n = j − 1, and t = k − 1 in (4).
Then, X
j
k=1
i − 1
k − 1
j − 1
j − k
=
i + j − 2
j − 1
= sij , the entries of the symmetric
Pascal matrix Sn. Hence, Sn = LnUn.
LINEAR ALGEBRA OF PASCAL MATRICES 4
This result can be used to determine the determinant of the symmetric
Pascal matrix, Sn.
Corollary 2.2. If Sn is the symmetric Pascal matrix of order n defined
by (1), then det(Sn) = 1, for any positive integer n.
Proof. Let Sn be the symmetric Pascal matrix of order n defined by
(1). By Theorem 2.1, we know that Sn = LnUn, where Ln is the lower
triangular Pascal matrix of order n defined by (2) and Un is the upper
triangular Pascal matrix of order n defined by (3). Since Ln and Un
are triangular matrices, then det(Ln) = 1 and det(Un) = 1. It follows
that det(Sn) = det(LnUn) = det(Ln)det(Un) = 1.
Definition 2.3. [5] Let A and B be n × n matrices. We say that
A is similar to B if there is an invertible n × n matrix P such that
P
−1AP = B.
Theorem 2.4. [1] Let Sn be the symmetric Pascal matrix of order n
defined by (1), then Sn is similar to its inverse S
−1
n
.
This result shows the following property of the eigenvalues of Sn.
Corollary 2.5. [1] Let Sn be the symmetric Pascal matrix of order
n defined by (1). Then the eigenvalues of Sn are pairs of reciprocal
numbers.
Proof. Let Sn be the symmetric Pascal matrix of order n defined by
(1) and λ be an eigenvalue of Sn. Since the det(Sn) = 1, we know Sn
is invertible. It follows that λ 6= 0 and, λ
−1
is an eigenvalue of S
−1
n
.
Since Sn and S
−1
n
are similar by Theorem 2.4, then Sn and S
−1
n have
the same eigenvalues. Hence, λ and λ
−1 are eigenvalues of Sn, and the
eigenvalues of Sn are pairs of reciprocal numbers.
Remark 1. If n is odd, since the eigenvalues must come in pairs, one
of the eigenvalues must be equal to 1.
Example 2.6. The eigenvalues of the symmetric Pascal matrix, S2,
are λ1 =
3 + √
5
2
and λ2 =
3 −
√
5
2
, where λ1λ2 = 1 gives a reciprocal
pair.
Example 2.7. For n odd, let n = 3. Then the eigenvalues of the symmetric
Pascal matrix, S3, are λ1 = 4+√
15, λ2 = 4−
√
15, and λ3 = 1.
We note that λ1λ2 = 1 gives a reciprocal pair and λ3 = 1 is a selfreciprocal.
LINEAR ALGEBRA OF PASCAL MATRICES 5
In their paper, A Note on Pascal’s Matrix, Cheon, Kim, and Yoon
found an interesting factorization of the lower triangular Pascal matrix,
Ln.
Theorem 2.8. [4] Let Gk =
In−k OT
O Sk
be a matrix of order n, where Sk is the matrix of order k defined by:
sij =
(
1 if i ≥ j
0 j>i
for every k = 1, 2, ..., n. Then the lower triangular Pascal matrix of
order n can be written as: Ln = GnGn−1 · · · G1.
For example,
L4 =
1 0 0 0 0
1 1 0 0 0
1 1 1 0 0
1 1 1 1 0
1 1 1 1 1
1 0 0 0 0
0 1 0 0 0
0 1 1 0 0
0 1 1 1 0
0 1 1 1 1
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
0 0 1 1 0
0 0 1 1 1
1 0 0 0 0
0 1 0 0 0
0 0 1 0 0
0 0 0 1 0
0 0 0 1 1
=
1 0 0 0 0
1 1 0 0 0
1 2 1 0 0
1 3 3 1 0
1 4 6 4 1
To further our studies of the lower triangular Pascal matrix, we are
interested in studying the inverse of this matrix.
Theorem 2.9. [6] Let Ln be the lower triangular Pascal matrix of order
n defined by (2), then
L
−1
n = ((−1)i−j
lij ).
Proof. We will show that Ln · L
−1
n = In.
By direct multiplication of Ln and L
−1
n we get the ij-th element of the
product:
Xn
k=1
(−1)k−j
liklkj . (5)
LINEAR ALGEBRA OF PASCAL MATRICES 6
If i < j then the element (5) is zero and if i = j, then the element (5)
is 1.
We will show that for i > j, the element (5) is zero.
If i > j, the element is:
i − 1
j − 1
j − 1
j − 1
−
i − 1
j
j
j − 1
+ ... + ( − 1)i − j
i − 1
i − 1
i − 1
j − 1
=
=
i − 1
j − 1
i − j
i − j
−
i − j
i − j − 1
+ ... + ( − 1)i − j
i − j
0
= 0.
Hence, Ln · L
−1
n = In and L
−1
n = ((−1)i−j
lij ) is the inverse of Ln.
There is another unique way in which the inverse of the lower triangular
Pascal matrix can be written, using the Hadamard product of matrices.
Definition 2.10. [8] Let A, B be m × n matrices. The Hadamard
product of A and B is defined by:
[A ◦ B]ij = [A]ij [B]ij , for 1 ≤ i ≤ m, 1 ≤ j ≤ n.
Theorem 2.11. [8] Let τn be a n × n lower triangular matrix defined
below as:
τij =
(
(−1)i−j
if i ≥ j ≥ 1,
0 otherwise
The inverse of the lower triangular matrix can be found using the
Hadamard product:
L
−1
n = Ln ◦ τn
For example, if n = 4, then:
L
−1
4 =
1 0 0 0
−1 1 0 0
1 −2 1 0
−1 3 −3 1
=
1 0 0 0
1 1 0 0
1 2 1 0
1 3 3 1
◦
1 0 0 0
−1 1 0 0
1 −1 1 0
−1 1 −1 1
3. Inverse of the Pascal Matrix Plus An Integer
In this section we are going to describe the inverse of Ln +kIn where
Ln is the lower triangular matrix of order n defined by (2), In is the
identity matrix and k is a positive integer. We call Ln +kIn the Pascal
matrix plus an integer. First, we are considering the case for k = 1. By
direct computation of the inverse of Ln+In, we can observe that there is
LINEAR ALGEBRA OF PASCAL MATRICES 7
a close relation between the inverse of Ln+In and the Pascal matrix Ln.
For example, for n = 4,
L4 + I4 =
2 0 0 0
1 2 0 0
1 2 2 0
1 3 3 2
,
(L4 + I4)
−1 =
1
2
0 0 0
−1
4
1
2
0 0
0
−1
2
1
2
0
1
8
0
−3
4
1
2
=
1 0 0 0
1 1 0 0
1 2 1 0
1 3 3 1
◦
1
2
0 0 0
−1
4
1
2
0 0
0
−1
2
1
2
0
1
8
0
−3
4
1
2
In their paper, Explicit Inverse of the Pascal Matrix Plus One, S.L.
Yang and Z.K. Liu showed that the inverse of Ln +In is the Hadamard
product between Ln and a lower triangular matrix. We are going to
describe this unique lower triangular matrix next. For this we need to
define the Euler polynomials.
Euler
พีชคณิตเชิงเส้นของเมทริกซ์ปาสกาลเยตส์ลินด์เซย์บทคัดย่อ สามเหลี่ยมปาสมีชื่อเสียงปรากฏในหลายพื้นที่คณิตศาสตร์ ทฤษฎีเลข คณิตศาสตร์เชิงการจัด และพีชคณิตมาจากสามเหลี่ยมนี้ทวินาม coeffi-เมทริกซ์ปาสกาลcients ซึ่งสร้างเมทริกซ์ง่าย ๆ ด้วยคุณสมบัติที่น่าสนใจเราสำรวจคุณสมบัติของเมทริกซ์นี้และค่าผกผันของการปาสกาลเมทริกซ์การบวกเมทริกซ์เอกลักษณ์ครั้งจำนวนเต็มบวกใด ๆ เราต่อไป พิจารณาเมทริกซ์เอกลักษณ์เรียกว่าเมตริกซ์สเตอร์ลิง ซึ่งสามารถ factorized ในเมตริกซ์ปาสกาล1. บทนำโบราณทางคณิตศาสตร์สามเหลี่ยม ปัจจุบันเรียกว่าสามเหลี่ยมปาสกาล เป็นอนันต์เลขตารางที่แสดงในแบบฟอร์มสามเหลี่ยม หมายเลขแสดงในรูปสามเหลี่ยมจะเรียกว่าสัมประสิทธิ์ทวินามnkซึ่งแทนจำนวนวิธีการรับนำผลที่ได้จาก n kไป แต่ละรายการในสามเหลี่ยมจะได้รับ โดยการเพิ่มเข้าด้วยกันรายการที่สองจากแถวด้านบน: ได้โดยตรงทางด้านซ้ายและหนึ่งตรงทางขวา รูปแบบนี้สามารถมองเห็นในภาพด้านล่างสามเหลี่ยมปาสกาลเป็นที่รู้จักในกว่าสิบศตวรรษ ชุดจำนวนที่สามเหลี่ยมปาสกาลถูกเรียกก่อน Blaiseวัน: 5 ธันวาคม 25571พีชคณิตเชิงเส้นของเมทริกซ์ปาสกาล 2ปาสกาล แม้ว่าเขาจะเกิดจาก ด้วยการเป็นคนแรกที่จะเผยแพร่ข้อมูลที่รู้จักสามเหลี่ยมในตำรับของเขา Trait´e du สามเหลี่ยมarithm´etique หมายเลขเดิมเกิดจากศึกษาที่อินเดียคณิตศาสตร์เชิงการจัดและดอกเบี้ยกรีกในหมายเลข figurate เหล่านี้หมายเลขถูกอย่างต่อเนื่องกล่าว โดย mathematicians อิสลามระหว่างคริสต์ศตวรรษ 10 และ ใน ศตวรรษ 11 โดยกวีเปอร์เซียชื่อOmar เคย์ยาม พวกเขายังได้เห็นในจีนในช่วงศตวรรษ 13สามเหลี่ยมปาสกาลถูกเผยแพร่อย่างเป็นทางในตำรับของปาสกาลหลังจากสิ้นพระชนม์ในค.ศ. 1665สามเหลี่ยมนี้เกิดขึ้นในหลายพื้นที่ของคณิตศาสตร์เช่นพีชคณิตความน่าเป็น และคณิตศาสตร์เชิงการจัด เรามีแรงจูงใจ โดยของปาสกาลสามเหลี่ยมความโดดเด่นในด้านคณิตศาสตร์และโปรแกรมประยุกต์หลายตัวในเฉพาะเมทริกซ์ปาสกาล เราต้องการไปศึกษาของเราพิจารณาคุณสมบัติต่าง ๆ และการเชื่อมต่อเฉพาะที่ปาสกาลเมทริกซ์มีการลำดับหมายเลขและฟังก์ชันอื่น ๆ2. เมทริกซ์ปาสกาลสามเหลี่ยมปาสกาลสามารถทับศัพท์เป็นเมตริกซ์ที่ประกอบด้วยสัมประสิทธิ์ทวินามเป็นองค์ประกอบ เราสามารถสร้างได้สามชนิดเมทริกซ์: สมมาตร ล่างสามเหลี่ยม และด้านบน สามเหลี่ยมใด ๆจำนวนเต็ม n > 0ปาสกาลเมทริกซ์สมมาตรของลำดับ n ถูกกำหนด โดย Sn = (sij), ที่sij =i + j − 2เจ− 1สำหรับ i, j = 1, 2,..., n (1)เราสามารถกำหนดล่างสามเหลี่ยมปาสกาลเมตริกซ์ของลำดับ n โดย Ln =(lij), ที่lij =(i−1j−1ถ้าผมเจ≥อื่น ๆ 0(2)กำหนด โดยสหประชาชาติด้านบนสามเหลี่ยมปาสกาลเมตริกซ์ของลำดับ n = (uij),ซึ่งuij =(j−1i−1ถ้า≥เจฉันอื่น ๆ 0(3)พีชคณิตเชิงเส้นของเมทริกซ์ปาสกาล 3เราสังเกตที่สหประชาชาติ = (Ln)Tสำหรับจำนวนเต็มบวก n ใด ๆตัวอย่างเช่น สำหรับ n = 5 ที่เรามี:S5 =1 1 1 1 11 2 3 4 51 3 6 10 151 4 10 20 351 5 15 35 70L5 =1 0 0 0 01 1 0 0 01 2 1 0 01 3 3 1 01 4 6 4 1U5 =1 1 1 1 10 1 2 3 40 0 1 3 60 0 0 1 40 0 0 0 1เมทริกซ์ปาสกาลเหล่านี้มีคุณสมบัติที่น่าสนใจบางอย่าง ซึ่งเรามีต่อไปทฤษฎีบท 2.1 [1] เป็นเมตริกซ์ปาสกาลสมมาตรของลำดับ n Sn ให้กำหนดตาม (1), Ln มีล่างสามเหลี่ยมปาสกาลเมตริกซ์ลำดับ n de-ปรับตาม (2), และสหประชาชาติอยู่บนสามเหลี่ยมปาสกาลเมตริกซ์ของลำดับ nกำหนดตาม (3), แล้ว Sn = LnUnหลักฐานการ Ln เป็นล่างสามเหลี่ยมปาสกาลเมตริกซ์ของ n ใบสั่งที่กำหนดให้ตาม (2) และสหประชาชาติอยู่บนสามเหลี่ยมปาสกาลเมตริกซ์ของ n ลำดับที่กำหนด(3) โดยการคูณของเมทริกซ์ Ln และสหประชาชาติโดยตรง เรารับการองค์ประกอบของ ij แค th ของผลิตภัณฑ์ LnUn:Xnk = 1likukj =Xnk = 1liklkj นับตั้งแต่สหประชาชาติ = (Ln)T.แล้ว Xnk = 1likljk =Xnk = 1ฉัน− 1k − 1เจ− 1k − 1=Xเจk = 1ฉัน− 1k − 1เจ− 1k − 1==Xเจk = 1ฉัน− 1k − 1เจ− 1เจ− kตั้งแต่บานธัท = 0 สำหรับ k > เจรหัสประจำตัว Vandermonde บอกว่า:Xnt = 0mtnn − t=m + nnสำหรับใด ๆ m, n, t ∈ N (4)ให้ m =ฉัน− 1, n = j − 1 และ t = k − 1 ใน (4)แล้ว Xเจk = 1ฉัน− 1k − 1เจ− 1เจ− k=i + j − 2เจ− 1= sij รายการของการสมมาตรปาสกาลเมตริกซ์ Sn. Hence, Sn = LnUn พีชคณิตเชิงเส้นของปาสกาลเมทริกซ์ 4สามารถใช้ผลนี้เพื่อกำหนดว่าดีเทอร์มิแนนต์ของการสมมาตรปาสกาลเมตริกซ์ SnCorollary 2.2 ถ้า Sn ปาสกาลเมทริกซ์สมมาตรของ n ลำดับที่กำหนด(1), แล้ว det(Sn) = 1 สำหรับจำนวนเต็มบวก n ใด ๆหลักฐานการ ให้ Sn เป็นเมตริกซ์ปาสกาลสมมาตรของ n ใบสั่งที่กำหนดโดย(1) โดยทฤษฎีบท 2.1 เรารู้ว่า Sn = LnUn, Ln อยู่ด้านล่างสามเหลี่ยมปาสกาลเมตริกซ์ n ใบสั่งที่กำหนดตาม (2) และสหประชาชาติอยู่บนสามเหลี่ยมปาสกาลเมตริกซ์ของ n ใบสั่งที่กำหนดตาม (3) Ln และสหประชาชาติเมทริกซ์สามเหลี่ยม แล้ว det(Ln) = 1 และ det(Un) = 1 เป็นไปตามที่ det(Sn) = det(LnUn) = det(Ln)det(Un) = 1ข้อกำหนดที่ 2.3 [5] ให้ A และ B เป็นเมทริกซ์ n n × เราบอกว่าคือคล้ายกับ B ถ้าเป็นเมตริกซ์ n × n สามารถหาอินเวอร์ส P ที่P−1AP = B.ทฤษฎีบทที่ 2.4 [1] เป็นเมตริกซ์ปาสกาลสมมาตรของลำดับ n Sn ให้กำหนดตาม (1), แล้ว Sn จะคล้ายการผกผันของ S−1n.ผลลัพธ์นี้แสดงคุณสมบัติต่อไปนี้ของเวกเตอร์ของ SnCorollary 2.5 [1] ให้ Sn เป็นเมตริกซ์ปาสกาลสมมาตรของใบสั่งn ที่กำหนดตาม (1) แล้ว เวกเตอร์ของ Sn คู่ซึ่งกันและกันหมายเลขหลักฐานการ ให้ Sn เป็นเมตริกซ์ปาสกาลสมมาตรของ n ใบสั่งที่กำหนดโดย(1) และλ eigenvalue ของ Sn ตั้งแต่ det(Sn) = 1 เรารู้ว่า Snจะสามารถหาอินเวอร์ส ตามที่ 6 = 0 λและ λ−1เป็นการ eigenvalue ของ S−1n.Since Sn and S−1nare similar by Theorem 2.4, then Sn and S−1n havethe same eigenvalues. Hence, λ and λ−1 are eigenvalues of Sn, and theeigenvalues of Sn are pairs of reciprocal numbers.Remark 1. If n is odd, since the eigenvalues must come in pairs, oneof the eigenvalues must be equal to 1.Example 2.6. The eigenvalues of the symmetric Pascal matrix, S2,are λ1 =3 + √52and λ2 =3 −√52, where λ1λ2 = 1 gives a reciprocalpair.Example 2.7. For n odd, let n = 3. Then the eigenvalues of the symmetricPascal matrix, S3, are λ1 = 4+√15, λ2 = 4−√15, and λ3 = 1.We note that λ1λ2 = 1 gives a reciprocal pair and λ3 = 1 is a selfreciprocal.LINEAR ALGEBRA OF PASCAL MATRICES 5In their paper, A Note on Pascal’s Matrix, Cheon, Kim, and Yoonfound an interesting factorization of the lower triangular Pascal matrix,Ln.Theorem 2.8. [4] Let Gk =In−k OTO Skbe a matrix of order n, where Sk is the matrix of order k defined by:sij =(1 if i ≥ j0 j>ifor every k = 1, 2, ..., n. Then the lower triangular Pascal matrix oforder n can be written as: Ln = GnGn−1 · · · G1.For example,L4 =1 0 0 0 01 1 0 0 01 1 1 0 01 1 1 1 01 1 1 1 11 0 0 0 00 1 0 0 00 1 1 0 00 1 1 1 00 1 1 1 11 0 0 0 00 1 0 0 00 0 1 0 00 0 1 1 00 0 1 1 11 0 0 0 00 1 0 0 00 0 1 0 00 0 0 1 0
0 0 0 1 1
=
1 0 0 0 0
1 1 0 0 0
1 2 1 0 0
1 3 3 1 0
1 4 6 4 1
To further our studies of the lower triangular Pascal matrix, we are
interested in studying the inverse of this matrix.
Theorem 2.9. [6] Let Ln be the lower triangular Pascal matrix of order
n defined by (2), then
L
−1
n = ((−1)i−j
lij ).
Proof. We will show that Ln · L
−1
n = In.
By direct multiplication of Ln and L
−1
n we get the ij-th element of the
product:
Xn
k=1
(−1)k−j
liklkj . (5)
LINEAR ALGEBRA OF PASCAL MATRICES 6
If i < j then the element (5) is zero and if i = j, then the element (5)
is 1.
We will show that for i > j, the element (5) is zero.
If i > j, the element is:
i − 1
j − 1
j − 1
j − 1
−
i − 1
j
j
j − 1
+ ... + ( − 1)i − j
i − 1
i − 1
i − 1
j − 1
=
=
i − 1
j − 1
i − j
i − j
−
i − j
i − j − 1
+ ... + ( − 1)i − j
i − j
0
= 0.
Hence, Ln · L
−1
n = In and L
−1
n = ((−1)i−j
lij ) is the inverse of Ln.
There is another unique way in which the inverse of the lower triangular
Pascal matrix can be written, using the Hadamard product of matrices.
Definition 2.10. [8] Let A, B be m × n matrices. The Hadamard
product of A and B is defined by:
[A ◦ B]ij = [A]ij [B]ij , for 1 ≤ i ≤ m, 1 ≤ j ≤ n.
Theorem 2.11. [8] Let τn be a n × n lower triangular matrix defined
below as:
τij =
(
(−1)i−j
if i ≥ j ≥ 1,
0 otherwise
The inverse of the lower triangular matrix can be found using the
Hadamard product:
L
−1
n = Ln ◦ τn
For example, if n = 4, then:
L
−1
4 =
1 0 0 0
−1 1 0 0
1 −2 1 0
−1 3 −3 1
=
1 0 0 0
1 1 0 0
1 2 1 0
1 3 3 1
◦
1 0 0 0
−1 1 0 0
1 −1 1 0
−1 1 −1 1
3. Inverse of the Pascal Matrix Plus An Integer
In this section we are going to describe the inverse of Ln +kIn where
Ln is the lower triangular matrix of order n defined by (2), In is the
identity matrix and k is a positive integer. We call Ln +kIn the Pascal
matrix plus an integer. First, we are considering the case for k = 1. By
direct computation of the inverse of Ln+In, we can observe that there is
LINEAR ALGEBRA OF PASCAL MATRICES 7
a close relation between the inverse of Ln+In and the Pascal matrix Ln.
For example, for n = 4,
L4 + I4 =
2 0 0 0
1 2 0 0
1 2 2 0
1 3 3 2
,
(L4 + I4)
−1 =
1
2
0 0 0
−1
4
1
2
0 0
0
−1
2
1
2
0
1
8
0
−3
4
1
2
=
1 0 0 0
1 1 0 0
1 2 1 0
1 3 3 1
◦
1
2
0 0 0
−1
4
1
2
0 0
0
−1
2
1
2
0
1
8
0
−3
4
1
2
In their paper, Explicit Inverse of the Pascal Matrix Plus One, S.L.
Yang and Z.K. Liu showed that the inverse of Ln +In is the Hadamard
product between Ln and a lower triangular matrix. We are going to
describe this unique lower triangular matrix next. For this we need to
define the Euler polynomials.
Euler
การแปล กรุณารอสักครู่..
