How to identify blocks in a diagonal block matrix?

8 visualizaciones (últimos 30 días)
Ludovic Tenorio
Ludovic Tenorio el 24 de Ag. de 2023
Editada: Stephen23 el 25 de Ag. de 2023
I have the following type of block diagonal matrix whose non-overlapping "blocks" I would like to indentify. These blocks may vary in size and may contain zeros.
M = [
1 1 1 0 0 0 0 0 0 0 0 0
1 1 1 0 0 0 0 0 0 0 0 0
1 1 1 0 0 0 0 0 0 0 0 0
0 0 0 1 1 1 0 0 0 0 0 0
0 0 0 1 1 1 0 0 0 0 0 0
0 0 0 1 1 1 0 0 0 0 0 0
0 0 0 0 0 0 1 0 0 0 0 0
0 0 0 0 0 0 0 1 0 1 0 0
0 0 0 0 0 0 0 0 1 1 1 1
0 0 0 0 0 0 0 1 1 1 0 1
0 0 0 0 0 0 0 0 1 0 1 1
0 0 0 0 0 0 0 0 1 1 1 1]
For this specific example, there are 4 distinct blocks whose indices are:
1, 2, 3 (3x3 block)
4, 5, 6 (3x3 block)
7 (1x1 block)
8, 9, 10, 11, 12 (4x4 block)
  2 comentarios
Paul
Paul el 25 de Ag. de 2023
What's the rule for identifying a distinct block? For example, why isn't 7 - 12 considered a single 5 x 5 block?
Ludovic Tenorio
Ludovic Tenorio el 25 de Ag. de 2023
The reason the element at indices (7,7) is its own block (and not part of the block on the lower right like you suggested), is because it's the only value over that row/colum (i.e. if you sum over either the 7th column or row, you get a value of one). I realize my definitition of block is a little arbitrary here, hopefully this clarifies things.

Iniciar sesión para comentar.

Respuesta aceptada

Stephen23
Stephen23 el 25 de Ag. de 2023
Editada: Stephen23 el 25 de Ag. de 2023
This is not very pretty, but it gets the job done. Note for simplicity it only handles square matrices and assumes square, non-overlapping blocks of data. Extending this approach to non-square matrices and blocks might be possible.
M = [1,1,1,0,0,0,0,0,0,0,0,0; 1,1,1,0,0,0,0,0,0,0,0,0; 1,1,1,0,0,0,0,0,0,0,0,0; 0,0,0,1,1,1,0,0,0,0,0,0; 0,0,0,1,1,1,0,0,0,0,0,0; 0,0,0,1,1,1,0,0,0,0,0,0; 0,0,0,0,0,0,1,0,0,0,0,0; 0,0,0,0,0,0,0,1,0,1,0,0; 0,0,0,0,0,0,0,0,1,1,1,1; 0,0,0,0,0,0,0,1,1,1,0,1; 0,0,0,0,0,0,0,0,1,0,1,1; 0,0,0,0,0,0,0,0,1,1,1,1];
char(M+'0') % just for compact display of the entire matrix
ans = 12×12 char array
'111000000000' '111000000000' '111000000000' '000111000000' '000111000000' '000111000000' '000000100000' '000000010100' '000000001111' '000000011101' '000000001011' '000000001111'
assert(isequal(diff(size(M)),0),'matrix must be square')
C = {};
B = 1; % block begin
E = B; % block end
R = size(M,1);
while B<R
while E<R && any(any(M(E+1:R,B:E)|M(B:E,E+1:R).'))
E = E+1;
end
C{end+1} = M(B:E,B:E); %#ok<SAGROW>
E = E+1;
B = E;
end
Checking:
C
C = 1×4 cell array
{3×3 double} {3×3 double} {[1]} {5×5 double}
C{:}
ans = 3×3
1 1 1 1 1 1 1 1 1
ans = 3×3
1 1 1 1 1 1 1 1 1
ans = 1
ans = 5×5
1 0 1 0 0 0 1 1 1 1 1 1 1 0 1 0 1 0 1 1 0 1 1 1 1

Más respuestas (0)

Categorías

Más información sobre Operating on Diagonal Matrices en Help Center y File Exchange.

Productos

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by