Minimum value in sub Matrix
4 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Zach Hudson
el 26 de Mayo de 2015
I have a 4 dimensional matrix that represents an array of arrays and I need to find the minimum value and its index in each of the sub arrays. IE I need to call min(example(:)) on example(:,:,1,1), example(:,:,1,2)... example(:,:,1,m), example(:,:,2,1)... example(:,:,2,m)... example(:,:,n,m). Any help with this would be much appreciated.
0 comentarios
Respuesta aceptada
Stephen23
el 26 de Mayo de 2015
Editada: Stephen23
el 26 de Mayo de 2015
X = [4D array];
Y = min(min(X,[],1),[],2);
Here is a simple example, with a 3D array:
>> X = reshape(1:24, [2,3,4])
X(:,:,1) =
1 3 5
2 4 6
X(:,:,2) =
7 9 11
8 10 12
X(:,:,3) =
13 15 17
14 16 18
X(:,:,4) =
19 21 23
20 22 24
>> Y = min(min(X,[],1),[],2)
Y(:,:,1) =
1
Y(:,:,2) =
7
Y(:,:,3) =
13
Y(:,:,4) =
19
You can see then it has calculated the minimum of the first two dimensions only, leaving the third dimension remaining. This works for 4D arrays too. If you do not want the leading scalar-dimensions to remain, simply use squeeze to remove them:
>> squeeze(Y)
ans =
1
7
13
19
2 comentarios
Stephen23
el 26 de Mayo de 2015
Editada: Stephen23
el 26 de Mayo de 2015
This is possible, but more complicated. Here is an example for a 3D array:
>> X = reshape(randperm(24), [2,3,4])
X(:,:,1) =
7 21 12
20 6 24
X(:,:,2) =
14 3 8
22 10 15
X(:,:,3) =
19 5 16
11 4 1
X(:,:,4) =
13 23 18
2 17 9
>> [Y,idy] = min(X,[],1);
>> [Z,idz] = min(Y,[],2);
>> N = size(X,3);
>> idy(sub2ind(size(idy),ones(N,1),squeeze(idz),(1:N)')) % row indices
ans =
2
1
2
2
>> squeeze(idz) % column indices
ans =
2
2
3
1
>> squeeze(Z) % the minimum values
ans =
6
3
1
2
This can be extended to work with 4D arrays... with some effort. It might be easier to perform this operation independently for each slice of the fourth dimension.
Más respuestas (0)
Ver también
Categorías
Más información sobre Deep Learning Toolbox 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!