如何获得矩阵的独特切片?

时间:2013-03-14 23:55:21

标签: matlab matrix language-features built-in

在matlab中,如果你有一个矩阵A,你可以找到包含B所有唯一行的矩阵A,如下所示:

B = unique(A,'rows');

我所拥有的是一个3d矩阵,其中行和列作为前两个维度,另外一个维度('切片')。

如何在矩阵A中获取包含所有唯一切片的3d矩阵?这是我想要的功能的一个例子:

>> A % print out A
A(:,:,1) =

    16     2     3    13
     5    11    10     8
     9     7     6    12
     4    14    15     1


A(:,:,2) =

     1     0     0     0
     0     1     0     0
     0     0     1     0
     0     0     0     1


A(:,:,3) =

    16     2     3    13
     5    11    10     8
     9     7     6    12
     4    14    15     1


A(:,:,4) =

     0     0     0     1
     0     0     1     0
     0     1     0     0
     1     0     0     0

>> unique(A,'slices'); % get unique slices

A(:,:,1) =

    16     2     3    13
     5    11    10     8
     9     7     6    12
     4    14    15     1


A(:,:,2) =

     1     0     0     0
     0     1     0     0
     0     0     1     0
     0     0     0     1


A(:,:,3) =

     0     0     0     1
     0     0     1     0
     0     1     0     0
     1     0     0     0

2 个答案:

答案 0 :(得分:6)

我将首先重塑A,以便每个切片成为一行(使用reshape命令)。然后使用unique(A, 'rows')。最后,将唯一的行重新整形为切片的相同形状。

例如:

% transforming so each row is a slice in row form
reshaped_A = reshape(A, [], size(A, 3))';

% getting unique rows
unique_rows = unique(reshaped_A, 'rows');

% reshaping back
unique_slices = reshape(unique_rows', size(A, 1), size(A, 2), []);

或全部在一行:

reshape(unique(reshape(A, [], size(A, 3))', 'rows')', size(A, 1), size(A, 2), [])

我没有检查上面这段代码所以请谨慎使用!但它应该提出这个想法。


编辑

这里正在处理你的数据(也修复了上面代码中的小错误):

>> reshaped_A = reshape(A, [], size(A, 3))'

reshaped_A =

Columns 1 through 11

16     5     9     4     2    11     7    14     3    10     6
 1     0     0     0     0     1     0     0     0     0     1
16     5     9     4     2    11     7    14     3    10     6
 0     0     0     1     0     0     1     0     0     1     0

Columns 12 through 16

15    13     8    12     1
 0     0     0     0     1
15    13     8    12     1
 0     1     0     0     0

这些^^行中的每一行都是原始切片之一

>> unique_rows = unique(reshaped_A, 'rows')

unique_rows =

Columns 1 through 11

 0     0     0     1     0     0     1     0     0     1     0
 1     0     0     0     0     1     0     0     0     0     1
16     5     9     4     2    11     7    14     3    10     6

Columns 12 through 16

 0     1     0     0     0
 0     0     0     0     1
15    13     8    12     1

这些^^是唯一的切片,但形状错误。

>> unique_slices = reshape(unique_rows', size(A, 1), size(A, 2), [])

unique_slices(:,:,1) =

 0     0     0     1
 0     0     1     0
 0     1     0     0
 1     0     0     0


unique_slices(:,:,2) =

 1     0     0     0
 0     1     0     0
 0     0     1     0
 0     0     0     1


unique_slices(:,:,3) =

16     2     3    13
 5    11    10     8
 9     7     6    12
 4    14    15     1

答案 1 :(得分:1)

一个非常简单且可扩展的解决方案是:

A = cat(3, [16 2 3 13;5 11 10 8;9 7 6 12;4 14 15 1], [1 0 0 0;0 1 0 0;0 0 1 0;0 0 0 1], [16 2 3 13;5 11 10 8;9 7 6 12;4 14 15 1], [0 0 0 1;0 0 1 0;0 1 0 0;1 0 0 0])
[n,m,p] = size(A);
a = reshape(A,n,[],1);
b = reshape(a(:),n*m,[])';
c = unique(b,'rows', 'stable')';    %If the 'stable' option is supported by your version.
%If the 'stable' option is not supported, but it's still required, use the index vector option, as required. 
%i.e.,
%[c,I,J] = unique(b,'rows');
unique_A = reshape(c,n,m,[])

<强>结果:

A(:,:,1) =

    16     2     3    13
     5    11    10     8
     9     7     6    12
     4    14    15     1


A(:,:,2) =

     1     0     0     0
     0     1     0     0
     0     0     1     0
     0     0     0     1


A(:,:,3) =

    16     2     3    13
     5    11    10     8
     9     7     6    12
     4    14    15     1


A(:,:,4) =

     0     0     0     1
     0     0     1     0
     0     1     0     0
     1     0     0     0


unique_A(:,:,1) =

     0     0     0     1
     0     0     1     0
     0     1     0     0
     1     0     0     0


unique_A(:,:,2) =

     1     0     0     0
     0     1     0     0
     0     0     1     0
     0     0     0     1


unique_A(:,:,3) =

    16     2     3    13
     5    11    10     8
     9     7     6    12
     4    14    15     1

来源:How to find unique pages in a 3d matrix?