Indecomposable and irreducible t-monomial matrices over commutative rings
We introduce the notion of the defining sequence of a permutation indecomposable monomial matrix over a commutative ring and obtain necessary conditions for such matrices to be indecomposable or irreducible in terms of this sequence.
Збережено в:
Дата: | 2016 |
---|---|
Автори: | , , , |
Формат: | Стаття |
Мова: | English |
Опубліковано: |
Інститут прикладної математики і механіки НАН України
2016
|
Назва видання: | Algebra and Discrete Mathematics |
Онлайн доступ: | http://dspace.nbuv.gov.ua/handle/123456789/155743 |
Теги: |
Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
|
Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
Цитувати: | Indecomposable and irreducible t-monomial matrices over commutative rings / V.M. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak // Algebra and Discrete Mathematics. — 2016. — Vol. 22, № 1. — С. 11-20. — Бібліогр.: 4 назв. — англ. |
Репозитарії
Digital Library of Periodicals of National Academy of Sciences of Ukraineid |
irk-123456789-155743 |
---|---|
record_format |
dspace |
spelling |
irk-123456789-1557432019-06-18T01:25:21Z Indecomposable and irreducible t-monomial matrices over commutative rings Bondarenko, V.M. Bortos, M. Dinis, R. Tylyshchak, A. We introduce the notion of the defining sequence of a permutation indecomposable monomial matrix over a commutative ring and obtain necessary conditions for such matrices to be indecomposable or irreducible in terms of this sequence. 2016 Article Indecomposable and irreducible t-monomial matrices over commutative rings / V.M. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak // Algebra and Discrete Mathematics. — 2016. — Vol. 22, № 1. — С. 11-20. — Бібліогр.: 4 назв. — англ. 1726-3255 2010 MSC:15B33, 15A30. http://dspace.nbuv.gov.ua/handle/123456789/155743 en Algebra and Discrete Mathematics Інститут прикладної математики і механіки НАН України |
institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
collection |
DSpace DC |
language |
English |
description |
We introduce the notion of the defining sequence of a permutation indecomposable monomial matrix over a commutative ring and obtain necessary conditions for such matrices to be indecomposable or irreducible in terms of this sequence. |
format |
Article |
author |
Bondarenko, V.M. Bortos, M. Dinis, R. Tylyshchak, A. |
spellingShingle |
Bondarenko, V.M. Bortos, M. Dinis, R. Tylyshchak, A. Indecomposable and irreducible t-monomial matrices over commutative rings Algebra and Discrete Mathematics |
author_facet |
Bondarenko, V.M. Bortos, M. Dinis, R. Tylyshchak, A. |
author_sort |
Bondarenko, V.M. |
title |
Indecomposable and irreducible t-monomial matrices over commutative rings |
title_short |
Indecomposable and irreducible t-monomial matrices over commutative rings |
title_full |
Indecomposable and irreducible t-monomial matrices over commutative rings |
title_fullStr |
Indecomposable and irreducible t-monomial matrices over commutative rings |
title_full_unstemmed |
Indecomposable and irreducible t-monomial matrices over commutative rings |
title_sort |
indecomposable and irreducible t-monomial matrices over commutative rings |
publisher |
Інститут прикладної математики і механіки НАН України |
publishDate |
2016 |
url |
http://dspace.nbuv.gov.ua/handle/123456789/155743 |
citation_txt |
Indecomposable and irreducible t-monomial matrices over commutative rings / V.M. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak // Algebra and Discrete Mathematics. — 2016. — Vol. 22, № 1. — С. 11-20. — Бібліогр.: 4 назв. — англ. |
series |
Algebra and Discrete Mathematics |
work_keys_str_mv |
AT bondarenkovm indecomposableandirreducibletmonomialmatricesovercommutativerings AT bortosm indecomposableandirreducibletmonomialmatricesovercommutativerings AT dinisr indecomposableandirreducibletmonomialmatricesovercommutativerings AT tylyshchaka indecomposableandirreducibletmonomialmatricesovercommutativerings |
first_indexed |
2025-07-14T07:59:05Z |
last_indexed |
2025-07-14T07:59:05Z |
_version_ |
1837608412124282880 |
fulltext |
Algebra and Discrete Mathematics RESEARCH ARTICLE
Volume 22 (2016). Number 1, pp. 11–20
© Journal “Algebra and Discrete Mathematics”
Indecomposable and irreducible
t-monomial matrices over commutative rings
Vitaliy M. Bondarenko, Maria Yu. Bortos,
Ruslana F. Dinis, Alexander A. Tylyshchak
Communicated by V. V. Kirichenko
Abstract. We introduce the notion of the defining sequence
of a permutation indecomposable monomial matrix over a commu-
tative ring and obtain necessary conditions for such matrices to be
indecomposable or irreducible in terms of this sequence.
Introduction
Let K be a commutative ring (with unity). By a monomial matrix
M = (mij) over K we mean a quadratic n × n matrix, in each row and
each column of which there is exactly one non-zero element. With such
matrix M one can associate the directed graph Γ(M) with n vertices
numbered from 1 to n and arrows i → j for all mij 6= 0. Obviously, Γ(M)
is the disjoint union of cycles, each of which has the same direction of
arrows. If there is only one cycle, the monomial matrix M is called cyclic
(in other words, it is a permutation indecomposable monomial matrix). A
cyclic matrix of the form
M =
0 . . . 0 m1n
m21 . . . 0 0
...
. . .
...
...
0 . . . mn,n−1 0
2010 MSC: 15B33, 15A30.
Key words and phrases: local ring, similarity, indecomposable matrix, irreducible
matrix, canonically t-cyclic matrix, defining sequence, group, representation.
12 Indecomposable and irreducible t-monomial matrices
we call canonically cyclic. The sequence
v = v(M) = (m21, . . . , mn−1,n, m1n)
we call the defining sequence of M , and write
M = M(v) = M(m21, . . . , mn−1,n, m1n).
The sequence v∗ = v∗(M) = (m1n, mn−1,n, . . . , m21) is called dual to
v and the matrix M∗ = M(v∗) dual to M .
When all elements mij of a monomial (respectively, cyclic or canonically
cyclic) matrix M are of the form tsij (t ∈ K), where sij > 0, the matrix
M is called t-monomial (respectively, t-cyclic or canonically t-cyclic);
obviously, then tsi 6= 0 for all i.
The most interesting cases are, obviously, those when the element t is
non-invertible.
Matrices of such form were studied by the authors in [1], and in this
paper we continue our investigation.
1. Defining sequences and indecomposability
Through this section K denotes a commutative local ring with maximal
ideal R = Rad K 6= 0 and t ∈ R. All matrices are considered over K. By
Es one denotes the identity s × s matrix.
1.1. Permutation similarity. Let
M =
0 . . . 0 tsn
ts1 . . . 0 0
...
. . .
...
...
0 . . . tsn−1 0
(∗)
be a canonically t-cyclic matrix. A permutation of the members xi = tsi
of v(M) of the form xi, xi+1, . . . , xm, x1, x2, . . . , xi−1, is called a cyclic
permutation. Two matrices M(v) and M(v′) is called cyclically similar if
v can be obtained from v′ by a such permutation.
It is easy to prove the following statement1.
Proposition 1. a) Two canonically t-cyclic matrices is permutation
similar if and only if they are cyclically similar.
b) The matrix transpose to a canonically t-cyclic matrix is permutation
similar to the dual one.
1This statement is valid for matrices with elements from any set if one modifies the
definitions accordingly.
V. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak 13
1.2. Conjecture. Let M(v) be a canonically t-cyclic n×n matrix with
defining sequence v = (x1, x2, . . . , xn), where xi = tsi (see (∗)).
The sequence v is called periodic with a period 0 < p < n if p|n and
xs+p = xs for any 1 6 s 6 n−p, and non-periodic if otherwise. In the case
of v to be periodic, the matrix M(v) can be reduced by a permutation of its
rows and column to the following block-monomial form: N = (Nij)
m
i,j=1,
m = n/p, where N21 = x2Em, N32 = x3Em, . . . , Nn,n−1 = xnEm and
N1m = x1M(1, 1, . . . , 1) with 1 to occur m times2. The m × m matrix
M(1, 1, . . . , 1) can be indecomposable or decomposable depending on
properties of the ring K, and therefore so can be the matrix M(v).
Conjecture 6 (V. M. Bondarenko, Private Communication). Any canon-
ically t-cyclic matrix over K with non-periodic defining sequence is inde-
composable.
It is obvious that the idea of a proof of this conjecture basing on
decomposition of a given matrix M into a direct sum of others ones, with
a final contradiction, is futile. It is most likely the best idea is to use the
simple fact that M is indecomposable if any idempotent matrix X such
that MX = XM is identity or zero. The main difficulty in this way is
that the non-periodicity of the defining sequence of M is not determined
by its local properties. In the next subsection we consider a special case
in which the condition of the non-periodicity is satisfied automatically.
1.3. 2-homogeneous defining sequences. Let M(v) be again a
canonically t-cyclic n×n matrix with defining sequence v =(x1, x2, . . . , xn),
where xi = tsi . The sequence v is said to be 2-homogeneous if it is trans-
lated by a cyclic permutation on that of the form (a, a, . . . , a, b, b, . . . , b),
where a and b 6= a both actually occur.
The following theorem proves Conjecture 1 for this case.
Theorem 1. Any canonically t-cyclic matrix over K with 2-homogeneous
defining sequence is indecomposable.
Proof. Let M = M(v) be a canonically t-cyclic n×n matrix and let v has
s coordinates equal to a = tp and the other ones equal to b = tq. Assume
without loss of generality that v = (a, a, . . . , a, b, b, . . . , b) and p < q. So
v = tpv0, where v0 = (1, 1, . . . , 1, tq−p, tq−p, . . . , tq−p). After replacing v
2To do it, one is to arrange the rows and columns in the order 2, 2+p, 2+2p,. . . , 2+(m-
1)p, 3, 3+p, 3+2p,. . . , 3+(m-1)p,. . . , p, 2p, 3p,. . . , mp, p+1, (p+1)+p, (p+1)+2p,. . . ,
(p+1)+(m-2)p,1.
14 Indecomposable and irreducible t-monomial matrices
by v0 and tq−p by t (again without loss of generality), we come to the
situation where
M =
0 . . . 0 0 . . . 0 t
1 . . . 0 0 . . . 0 0
...
. . .
...
...
. . .
...
...
0 . . . 1 0 . . . 0 0
0 . . . 0 t . . . 0 0
...
. . .
...
...
. . .
...
...
0 . . . 0 0 . . . t 0
with s elements to equal 1 and k = n − s elements equal to t.
Arrange the rows and columns of the matrix M in the order 1, 2, . . . ,
n − k, n, n − 1, . . . , n − k + 2, n − k + 1, denoting the new matrix by N :
N =
(
N11 N12
N21 N22
)
=
0 . . . 0 0 t 0 . . . 0
1 . . . 0 0 0 0 . . . 0
...
. . .
...
...
...
...
. . .
...
0 . . . 1 0 0 0 . . . 0
0 . . . 0 0 0 t . . . 0
...
. . .
...
...
...
...
. . .
...
0 . . . 0 0 0 0 . . . t
0 . . . 0 1 0 0 . . . 0
.
Prove that there is not a non-trivial idempotent matrix commuting
with N (see the previous subsection).
Let C be an n × n matrix such that NC = CN , where
C =
(
C11 C12
C21 C22
)
=
=
c11 . . . c1,n−k c1,n−k+1 . . . c1n
...
. . .
...
...
. . .
...
cn−k,1 . . . cn−k,n−k cn−k,n−k+1 . . . cn−k,n
cn−k+1,1 . . . cn−k+1,n−k cn−k+1,n−k+1 . . . cn−k+,n
...
. . .
...
...
. . .
...
cn1 . . . cn,n−k cn,n−k+1 . . . cnn
with the same partition as N .
V. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak 15
We denote by (i, j) the scalar equality (NC)ij = (CN)ij . All compa-
risons below are considered modulo the maximal ideal R = Rad K (that
is, they are equalities over the residue field K/R). We use by default the
following simple fact: tx = ty implies x ≡ y (x, y ∈ K).
Since every jth column of CN with j > n − k consists of elements
from tK, we have from the equations (i, j) for i = 2, 3, . . . , n − k and
i = n that
C ≡
(
C11 0
C21 C22
)
=
=
c11 . . . c1,n−k 0 . . . 0
...
. . .
...
...
. . .
...
cn−k,1 . . . cn−k,n−k 0 . . . 0
cn−k+1,1 . . . cn−k+1,n−k cn−k+1,n−k+1 . . . cn−k+1,n
...
. . .
...
...
. . .
...
cn1 . . . cn,n−k cn,n−k+1 . . . cnn
.
The equations (i, j) for i, j = 1, 2, . . . , n − k mean that the matrix
C11 commutes modulo R with the lower Jordan block N11 and hence the
matrix C11 is lower triangular modulo R (see, e. g., [2, Chap. VIII] or
[3, Theorem 3.2.4.2]).
Further, from the equlities
(n − k + i, n − k + j + 1) : tCn−k+i+1,n−k+j+1 = tCn−k+i,n−k+j
for 1 6 i < j 6 k − 1 it follows that all elements of the matrix C22
belonging to its lth upper diagonal3 are pairwise comparable modulo R,
1 6 l 6 k − 2. But since the equalities
(n − k + i, n − k) : tCn−k+i+1,n−k = Cn−k+i,n,
1 6 i 6 k − 1, imply that the last elements of all (mentioned above) upper
diagonals are comparable with 0, we have eventually that the matrix C22
is, as well as C11 (see above), upper triangular modulo R.
Finally, from the equalities
I. (2, 1) : c11 = c22, (3, 2) : c22 = c33, . . .,
(n − k, n − k − 1) : cn−k−1,n−k−1 = cn−k,n−k,
II. (n − k + 1, n − k + 2) : tcn−k+1,n−k+1 = tcn−k+2,n−k+2,
(n − k, n − k + 1) : tcn−k,n−k = tcn−k+1,n−k+1, . . . ,
(n − 1, n) : tcn−1,n−1 = tcn,n,
3The lth upper diagonal of a matrix M = (mij), where l > 1, is the collection of
elements mi,i+l.
16 Indecomposable and irreducible t-monomial matrices
III. (n, n − k) : cn−k,n−k = cn,n,
it follows that c11 ≡ c22 ≡ · · · ≡ cnn.
Thus we prove that the matrix C is comparable to an upper triangular
one with the same elements on the main diagonal. It easily follows that
if C2 ≡ C, then C ≡ En or C ≡ 0, and, consequently, because the
comparisons are modulo the only maximal ideal of K, C = En or C = 0,
respectively.
1.4. Applications in the representation theory of groups.
Through this subsection K is as above and of characteristic ps (p is
simple, s > 1). All groups G are assumed to be finite of order |G| > 1.
The number of nonequivalent indecomposable matrix K-representations
of degree n of a group G is denoted by indK(G, n).
From [4] it follows that indK(G, n) > |K/R| for any p-group G of
order |G| > 2 and n > 1. Here we strengthen this result in the case of
both cyclic groups and radicals.
Theorem 2. Let R = tK 6= 0 with t being nilpotent. Then, for a cyclic
p-group G of some order N (hence of greater order), indK(G, n) > n − 1
for any n > 1.
Proof. Let S be an n × n matrix over K that is nilpotent modulo R; then
Sn ≡ 0 (mod R) and S2n ≡ 0 (mod R2). It is easy to see that the map
ΓS : a → ΓS(a) = En + uS with u = tm−2 is a K-representation of a
cyclic group G = 〈a〉 of an order pr > 2n, r > 2. It is indecomposable
if and only if so is modulo Ann u := {x ∈ K | tm−2x = 0} = R2, and
representations ΓS and ΓS′ are equivalent if and only if the matrices S
and S′ are similar modulo R2.
Consider the K-representations ΓMk
, k = 1, 2, . . . , n − 1, with the
matrices Mk = M(1, . . . , 1, t . . . , t), where 1 occurs k times (see the
previous subsection). They are non-equivalent, because Mk has rank
k modulo R, and indecomposable by Theorem 1 (taking into account
that Rad (K/Ann u) = R/R2 is a principal ideal of K/R2 generated by
t + R2 6= R2).
Theorem 3. Let the characteristic of K is p and R = tK 6= 0 with t2 = 0.
Then, for any cyclic p-group G and n > |G|, indK(G, n) > |G| − 2.
Proof. We use the notation of the proof of Theorem 2. It is easy to
see that, for 0 < k < n, the map Λk : a → Λ(a) = En + Mk is a K-
representation of a cyclic group G = 〈a〉 if k + 2 6 |G|; in particular, if
0 < k < |G| − 1 6 n − 1. By the last condition, their number is equal to
V. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak 17
|G| − 2. The representations Λk are indecomposable by Theorem 1, and
non-equivalent (see the previous proof).
2. Defining sequences and irreducibility
Through this section K also denotes a commutative local ring with
maximal ideal R = Rad K 6= 0; t 6= 0 is any element of R unless otherwise
stated.
For an n × n matrix A over K, we denote by [i
a
→ j]+ (resp. [i
a
→ j]−)
the following similarity transformation of A: adding ith row (resp. column),
multiplied by a, to jth row (resp. column), and then subtracting jth
column (resp. rows), multiplied by a, from ith column (resp. row).
2.1. Theorems on irreducible canonically t-cyclic matrices.
Let M = M(v) be a canonically t-cyclic n × n matrix with defining
sequence v = (x1, x2, . . . , xn), where xi = tsi (see (∗)). We call the
sequence v0 = v0(M) = (s1, s2, . . . , xn) the weighted sequence of M , and
the number w = w(M) = s1 + s2 + · · · + sn the weight of M4.
From the results of [1] it follows the next theorem5.
Theorem 4. If the matrix M(v) is irreducible, then its weight is prime
to n.
Corollary 1. If t2 = 0 and the matrix M(v) is irreducible, then its rank
modulo R is prime to n.
By a connected subsequence of the length 1 6 l 6 n of a defining
sequence v = (x1, x2, . . . , xn) we mean any subsequence which maps by a
cyclic permutation on one of the form x1, x2, . . . , xl. If it is 2-homogeneous
i. e., by analogy with the above said, has the form
u = (ti, ti, . . . , ti, tj , tj , . . . , tj), ti 6= tj , where ti and tj both actually
occur, then the pair (p, q), consisting of the numbers of occurrences of ti
and tj we call the type of u.
We shall obtain the following theorem as a consequence of statements
in more general situations.
Theorem 5. If the matrix M(v) with 2-homogeneous defining sequence
v is irreducible and n > 5, t2 = 0, then its weight is equal 1, 2 or n − 1.
Concerning the cases when the weight is equal 1, n − 1 see [1, Intro-
duction].
4One can write M(t, v0) instead of M(v) (as in [1]).
5A quadratic matrix is irreducible if it is not similar to a 2×2 upper block triangular
matrix with quadratic diagonal blocks.
18 Indecomposable and irreducible t-monomial matrices
2.2. Defining sequences with subsequences of type (2, 4)
and (4, 2). We are interested in the cases when a 2-homogeneous subse-
quence has the form (ts, ts, ts, ts, 1, 1) or (1, 1, ts, ts, ts, ts). Since they are
mutually dual (see Proposition 1), we consider only the first case.
Proposition 2. If tm = 0 and a canonically t-cyclic n × n matrix M(v)
is irreducible, then the sequence v does not contain a subsequence of the
form (ts, ts, ts, ts, 1, 1) with m 6 2s < 2m.
We prove a more general statement replacing (ts, ts, ts, ts, 1, 1) by
(ti, tj , tp, tq, 1, 1) with 0 < i, j, p, q < m, i + j > m, p + q > m, assuming
(by Proposition 1) that the subsequence is the beginning of v and (by
Theorem 4) that n > 6.
This follows from the following: if we perform with the reducible
matrix
N =
0 0 . . . 0 0 −tj 1 0 0
1 0 . . . 0 0 0 0 0 0
0 α1 . . . 0 0 0 0 0 0
...
...
. . .
...
...
...
...
...
...
0 0 . . . αk 0 0 0 0 0
0 0 . . . 0 ti 0 0 0 0
0 0 . . . 0 0 0 0 0 tq
0 0 . . . 0 0 0 1 0 0
0 0 . . . 0 0 0 0 tp 0
(k = n − 6) the transformation [(n − 2)
tj
→ (n − 3)]−, [1
−1
→ (n − 1)]+,
[2
−tp
→ n]+, and arrange the rows and columns of the resulting monomial
matrix
N ′ =
0 0 . . . 0 0 0 1 0 0
1 0 . . . 0 0 0 0 0 0
0 α1 . . . 0 0 0 0 0 0
...
...
. . .
...
...
...
...
...
...
0 0 . . . αk 0 0 0 0 0
0 0 . . . 0 ti 0 0 0 0
0 0 . . . 0 0 0 0 0 tq
0 0 . . . 0 0 tj 0 0 0
0 0 . . . 0 0 0 0 tp 0
in the order n − 4, n − 3, n − 1, n, n − 2, 1, 2, . . . , n − 5, we get the matrix
M(ti, tj , tp, tq, 1, 1, α1, . . . , αk).
V. Bondarenko, M. Bortos, R. Dinis, A. Tylyshchak 19
2.3. Defining sequences with subsequences of type (3, 3). We
consider the cases, when a 2-homogeneous subsequence has the form
(ts, ts, ts, 1, 1, 1) or (1, 1, 1, ts, ts, ts), in the same way as those in subsection
2.2; therefore we present only the main part and do not repeat similar
assumptions and comments.
Proposition 3. If tm = 0 and a canonically t-cyclic n × n matrix M(v)
is irreducible, then the sequence v does not contain a subsequence of the
form (ts, ts, ts, 1, 1, 1) with m 6 2s < 2m.
We prove a more general statement replacing (ts, ts, ts, 1, 1, 1) by
(ti, tj , tp, 1, 1, 1) with 0 < p, q, j < m, i + j > m, 2p > m.
This follows from the following: if we perform with the reducible
matrix
N =
0 0 0 . . . 0 0 −tj 1 0
1 0 −tp . . . 0 0 0 0 0
0 1 0 . . . 0 0 0 0 0
0 0 α1 . . . 0 0 0 0 0
...
...
...
. . .
...
...
...
...
...
0 0 0 . . . αk 0 0 0 0
0 0 0 . . . 0 ti 0 0 0
0 0 0 . . . 0 0 0 0 tp
0 0 0 . . . 0 0 0 1 0
(k = n − 6) the transformation [(n − 1)
tj
→ (n − 2)]−, [1
−1
→ n]+,
[2
−tp
→ (n − 1)]+, [3
−tp
→ 1]+, and arrange the rows and columns of the
resulting monomial matrix
N ′ =
0 0 0 . . . 0 0 0 1 0
1 0 0 . . . 0 0 0 0 0
0 1 0 . . . 0 0 0 0 0
0 0 α1 . . . 0 0 0 0 0
...
...
...
. . .
...
...
...
...
...
0 0 0 . . . αk 0 0 0 0
0 0 0 . . . 0 ti 0 0 0
0 0 0 . . . 0 0 0 0 tp
0 0 0 . . . 0 0 tj 0 0
in the order n − 3, n − 2, n, n − 1, 1, 2, . . . , n − 4, we get the matrix
M(ti, tj , tp, 1, 1, 1, α1, α2, . . . , αk).
20 Indecomposable and irreducible t-monomial matrices
2.4. Proof of Theorem 5. The proof follows from Propositions 2
and 3. Indeed, the first proposition implies at once that M = M(v) is
reducible, if w(M) = n−2, and the second one that M = M(v) is reducible,
if 2 < w(M) < n − 2 (in both the cases it need to take m = 2, s = 1).
In conclusion, we note that in the cases n = 2, 3 the theorem is trivial,
in the case n = 4 it follows from Theorem 4 and in the case n = 5 there
is the only exception, namely the matrix M(1, 1, t, t, t, ) of weight 3 is
irreducible.
The theorem is proved.
References
[1] V. M Bondarenko, M. Yu. Bortos, M. Yu. Dinis, A. A. Tylyshchak, Reducibility
and irreducibility of monomial matrices over commutative rings, Algebra Discrete
Math. 16 (2013), no. 2, pp. 171–187.
[2] F. R. Gantmaher, The theory of matrices, FIZMATLIT, 2004, 560 p.
[3] R. A. Horn C. R. Johnson, Matrix Analysis, 2nd ed., Cambridge University Press,
Cambridge, 2013, 643 p.
[4] P. M. Gudivok, I. B. Chukhraj, On the number of indecomposable matrix rep-
resentations of given degree of a finite p-group over commutative local rings of
characteristic ps, Nauk. Visn. Uzhgorod Univ. Ser. Mat. Inform. 5 (2000), pp.
33–40 (in Ukrainian).
Contact information
V. M. Bondarenko,
M. Yu. Bortos
Institute of Mathematics, Tereshchenkivska str.,
3, 01601 Kyiv, Ukraine
E-Mail(s): vit-bond@imath.kiev.ua,
bortosmaria@gmail.com
Web-page(s): http://www.imath.kiev.ua
R. F. Dinis Faculty of Mechanics and Mathematics, Kyiv Na-
tional Taras Shevchenko Univ., Volodymyrska
str., 64, 01033 Kyiv, Ukraine
E-Mail(s): ruslanadinis@ukr.net
A. A. Tylyshchak Faculty of Mathematics, Uzhgorod National
Univ., Universytetsyka str., 14, 88000 Uzhgorod,
Ukraine
E-Mail(s): alxtlk@gmail.com
Received by the editors: 29.08.2016.
|