ÓÄÊ 512.643.8
The Moore - Penrose Inverse for a Certain Class of Block Matrices
(Submitted by academician A. B. Nersessyan 10/III 2005)
1. The term semi-magic square
is attributed to an n×n real matrix having the sum of elements in
each row and each column equal to an identical constant (see [1], for instance).
As was shown in [2] , the Moore - Penrose inverse of a semi-magic square is
semi-magic as well. In this paper we introduce more general classes of matrices
(see Definitions 1.1 and 2.1 below) and establish a similar property of their
Moore - Penrose inverse. is a
magic rectangle with r = 6 and
c = 4. ◁ Basing on the relation (1.4), we can use another notation for the set of
magic rectangles with fixed row sum and column sum. Namely, simultaneously with
the notation MR(m, n; r, c) we will use a notation MR[m, n : g] which implies that r = ng and
c = mg. with submatrices
Aij Î
MR(mi, nj; rij, cij)
(or Aij Î
MR[mi, nj : gij] , in
another notation), where i = 1,2,...,p , j = 1,2,...,q and
= m, = n,
will be referred to as block magic rectangle. Insert diagonal matrices we will denote the set of block magic rectangles partitioned into blocks
correspondingly to (2.2) with row sums and column sums defined by the matrices
(2.3).
Yerevan State University
Let
Rn be the space of real n-dimensional column
vectors and Rm×n be the space of real
m×n matrices.
Definition 1.1. A matrix A = [aij] Î Rm×n is referred to as magic
rectangle if there exist constants r and c such that
the sum of elements in each row and each column is equal to r and
c , respectively, i.e.
n
å
j=1 aij = r ,
i = 1,2,...,m , (1.1)
We
will call the constants r and c row sum and
column sum, respectively.
m
å
i=1 aij = c ,
j = 1,2,...,n . (1.2)
Example 1.1. The matrix
A =
é
ê
ë
1
0
5
3
4
-1 ù
ú
û
Let us
denote by MR(m, n) the set of m × n magic rectangles. It can be readily shown that
MR(m,n) is a subspace of Rm×n. Next, MR(m, n; r, c) will denote
the set of m×n magic rectangles with row sum r and column sum c. Let
en = [11...1]T be the n-dimensional column vectors of ones.
Then the properties (1.1) and (1.2) can be written, respectively, as follows:
It
can be easily found the following relation between row sum, column sum and the
size of the matrix:
Aen = rem, ATem = c
en. (1.3)
mr = nc. (1.4)
Obviously, for
m = n a magic rectangle becomes a semi-magic square.
2. Let us define a class of block matrices composed of magic
rectangles.
Definition
2.1 A matrix A = [aij] Î
Rm×n represented in the block form
A =
é
ê
ê
ê
ê
ë
A11
A12
...
A1q
A21
A22
...
A2q
...
...
...
...
Ap1
Ap2
...
Apq
ù
ú
ú
ú
ú
û (2.1)
and p × q matrices
M =
é
ê
ê
ê
ê
ê
ë
m1
m2
0
0
···
mp
ù
ú
ú
ú
ú
ú
û Î
Rp×p , N =
é
ê
ê
ê
ê
ê
ë
n1
n2
0
0
···
nq
ù
ú
ú
ú
ú
ú
û Î
Rq×q (2.2)
into our consideration. By
R =
é
ê
ê
ê
ê
ë
r11
r12
...
r1q
r21
r22
...
r2q
...
...
...
...
rp1
rp2
...
rpq
ù
ú
ú
ú
ú
û , C =
é
ê
ê
ê
ê
ë
c11
c12
...
c1q
c21
c22
...
c2q
...
...
...
...
cp1
cp2
...
cpq
ù
ú
ú
ú
ú
û (2.3)
BMR(M, N ; R , C) (2.4)
In accordance with relation (1.4) and
notation accepted in the previous section, we have
for i = 1,2,...,p and
j = 1,2,...,q . Consider a matrix
rij = gijnj ,
cij = gijmi (2.5)
Then the relations (2.5) can be
written in the matrix form:
G =
é
ê
ê
ê
ê
ë
g11
g12
...
g1q
g21
g22
...
g2q
...
...
...
...
gp1
gp2
...
gpq
ù
ú
ú
ú
ú
û . (2.6)
Therefore, we will also use another notation for the set of block magic
rectangles (2.4), that is
R = GN , C = MG . (2.7)
The properties (1.3) for the blocks of matrix (2.1) look as
BMR[M,N : G] . (2.8)
where i = 1,2,...,p and
j = 1,2,...,q . Let us define block matrices
Aijenj = rijemi ,
AijTemi = cijenj , (2.9)
with
blocks of the size mi × 1 , i = 1,2,...,p (in the
matrix Emp) and nj × 1 ,
j = 1,2,...,q (in the matrix Enq). By a
straightforward verification it can be easily shown that relations (2.9) are
equivalent to the following ones:
Emp =
é
ê
ê
ê
ê
ê
ë
em1
0
...
0
0
em2
...
0
...
...
...
...
0
0
...
emp
ù
ú
ú
ú
ú
ú
û Î
Rm×p , Enq =
é
ê
ê
ê
ê
ê
ë
en1
0
...
0
0
en2
...
0
...
...
...
...
0
0
...
enq
ù
ú
ú
ú
ú
ú
û Î
Rn×q
3. Remind, that the Moore-Penrose inverse A+
Î Rn×m of a matrix A Î Rm×n is uniquely determined by the
properties
AEnq = EmpR ,
ATEmp = EnqCT .
(see
[3] , for instance).
a)
AA+A = A ,
b)
A+AA+ = A+ ,
c)
(A+A)T = A+A ,
d)
(AA+)T = AA+
The main result of
the paper is formulated as follows.
Theorem 3.1 If A Î BMR(M, N ; R , C) then
A+ Î
BMR(N, M ; ) ,
where
= N-1/2(M1/2RN-1/2)+M1/2 , (3.1)
Let us give an equivalent formulation of Theorem 3.1 , connected
with the notation of type (2.8) for a set of block magic
rectangles.
= N1/2(M-1/2CN1/2)+M-1/2 . (3.2)
Theorem 3.1A If A Î BMR[M, N : G] then A+ Î
BMR[N, M :
] ,
where
Example 3.1. Consider a block magic rectangle
= N-1/2(M1/2GN1/2)+M-1/2 . (3.3)
According
to (2.2),(2.3),(2.5),(2.6), we have
A =
é
ê
ê
ê
ê
ê
ê
ë
1
0
5
1
1
0
3
4
-1
1
1
0
1
0
2
4
2
2
1
2
0
1
5
2
1
0
2
3
3
2
1
2
0
4
2
2
ù
ú
ú
ú
ú
ú
ú
û
Î
R6×6
(p = 2 , q = 3) .
and
M =
é
ê
ë
2
0
0
4
ù
ú
û
, N =
é
ê
ê
ë
3
0
0
0
2
0
0
0
1 ù
ú
ú
û
The Moore-Penrose inverse of the matrix A calculated using
MATLAB is
R =
é
ê
ë
6
2
0
3
6
2
ù
ú
û
, C =
é
ê
ë
4
2
0
4
12
8
ù
ú
û
, G =
é
ê
ë
2
1
0
1
3
2
ù
ú
û
.
For this
matrix
A+ =
é
ê
ê
ê
ê
ê
ê
ë
-0.2411
0.4256
0.3212
-0.1788
0.4879
-0.6788
0.2589
-0.0744
-0.2788
0.1212
-0.4121
0.5212
0.2589
-0.0744
-0.0788
0.0212
-0.1121
0.1212
-0.0267
-0.0267
0.0864
-0.1136
-0.0136
0.1864
-0.0267
-0.0267
-0.0136
0.1864
0.0864
-0.1136
-0.0583
-0.0583
0.0340
0.0340
0.0340
0.0340
ù
ú
ú
ú
ú
ú
ú
û
.
=
é
ê
ê
ê
ë
0.1845
-0.0485
-0.0534
0.1456
-0.1165
0.1359
ù
ú
ú
ú
û
,
=
é
ê
ê
ê
ë
0.2767
-0.0364
-0.0534
0.0728
-0.0583
0.0340
ù
ú
ú
ú
û
,
It is of interest to examine some particular cases of block magic
rectangles for which the general formula, describing the Moore-Penrose inverse,
is considerably simplified.
=
é
ê
ê
ê
ë
0.0922
-0.0121
-0.0267
0.0364
-0.0583
0.0340
ù
ú
ú
ú
û
.
Case 1. Suppose that in block representation (2.1)
p = q = 1 , i.e. let A Î
MR(m, n ; r, c) . Then, by formulae (3.1) and (3.2), we get
Recall that for a scalar a , which can be considered as
1 × 1 matrix, a+ = 1/a , if a ¹ 0 , and a+ = 0 , if
a = 0 . So, we arrive at the following
statement.
= n-1/2(m1/2rn-1/2)+m1/2
= n-1/2n1/2r+m-1/2m1/2 = r+ ,
= n1/2(m-1/2c n1/2)+m-1/2 =
n1/2n-1/2c+m1/2m-1/2 = c+ .
Theorem 3.2 If A Î
MR(m, n ; r , c) then A+ Î
MR(n, m ; r+ , c+) .
Note, that the last proposition has been obtained recently in
[4].
We can also give an equivalent
formulation of Theorem 3.2 . Let A Î
MR[m, n : g] . According to (3.3), we find
Thus, the result can be stated as follows.
=n-1/2(m1/2gn1/2)+m-1/2 = n-1/2n-1/2g+m-1/2m-1/2 =
1
mn g+ .
Theorem 3.2A
If A Î MR[m, n : g] then
A+ Î MR[n, m : (mn)-1 g+] .
Case 2. Let in block form (2.1)
m1 = m2 = ¼ = mp
º k and n1 = n2 = ¼ = nq º l . It is
clear, that k = m/p and l = n/q . Thereby, according to (2.2),
we have M = kIp and N = lIq where
Ip and Iq are identity matrices of
pth and qth order, respectively. In this case our formulae (3.1) and
(3.2) become extremely simple, i.e.
Thus,
we get
Theorem 3.3 If A Î
BMR(kIp, lIq ; R , C) then
A+ Î
BMR(lIq, kIp ; R+ , C+) .
Next, from (3.3) we obtain
Consequently, the result may be formulated as
follows.
Theorem
3.3A If A Î
BMR[kIp, lIq : G] then
A+ Î
BMR[lIq, kIp : (kl)-1 G+] .
1. Weiner L. M. - Amer. Math.
Monthly 1955. V. 62. P. 237-239.
2. Schmidt
K., Trenkler G. - Int. J. Math. Educ. Sci. Technol. 2001. V. 32.
No. 4. P. 624-629.
3. Strang G. Linear Algebra and its Applications. Academic Press. 1976.
4. Hakopian Yu. R, Eloyan A. N.,
Khachatryan D. E. - Algebra, Geometry & their Applications.
Seminar Proceedings. Yerevan State University. 2004. V. 3-4. P. 30-34.