How can I create a unique matrix by removing rows with similar elements?

A = [11 12 1 3

8 9 10 12

227 228 217 219

217 218 8 7

224 225 226 228

443 444 433 435

433 434 224 223

440 441 442 444

659 660 649 651

649 650 440 439

656 657 658 660

875 876 865 867

865 866 656 655

872 873 874 876

1091 1092 1081 1083

1081 1082 872 871

1088 1089 1090 1092

1307 1308 1297 1299

1297 1298 1088 1087

1304 1305 1306 1308

1523 1524 1513 1515

1513 1514 1304 1303

1520 1521 1522 1524

3 1 12 11

12 10 9 8

6 4 3 2

9 7 6 5

219 217 228 227

228 226 225 224

222 220 219 218]

In A, I want to delete rows with similar elements. Because the elements in rows 1 and 3 are the same as those in rows 24 and 28, I want to delete 24 and 28. The output should look like this:

A_new = [[11 12 1 3

8 9 10 12

227 228 217 219

217 218 8 7

224 225 226 228

443 444 433 435

433 434 224 223

440 441 442 444

659 660 649 651

649 650 440 439

656 657 658 660

875 876 865 867

865 866 656 655

872 873 874 876

1091 1092 1081 1083

1081 1082 872 871

1088 1089 1090 1092

1307 1308 1297 1299

1297 1298 1088 1087

1304 1305 1306 1308

1523 1524 1513 1515

1513 1514 1304 1303

1520 1521 1522 1524

12 10 9 8

6 4 3 2

9 7 6 5

228 226 225 224

222 220 219 218]

Answer 1

[Prakash S R]

Hi,

You sort the rows so they are order-independent, then you can use unique() with the 'rows' option. However, the results are sorted along the row axis, which you will need to undo to get the result you want:

[B, Bi] = unique(sort(A,2), 'rows');
A_new = A(Bi,:);

-Prakash

[GANESH R]

This works for me. Thanks

Answer 2

[ClementJ]

Hi,

Maybe, you can use:

%% load A data
load('A.mat')

%% process
sum_A = sum(A,2);
[~,ia,~] = unique(sum_A);

%% change A value
A_news = A(ia,:);

[Akira Agata]

+1

To make it more robust and accurate, please modify a litte bit, like:

sort_A = sort(A, 2);
[~, ia] = unique(sort_A, 'stable', 'rows');

A_new = A(ia, :);

[GANESH R]

This isn't working.

[Akira Agata]

Seems strange... It should work.

Let me check by small example:

% Small example
A = [...
    1 2 3;...
    4 6 5;...
    1 1 1;...
    1 3 2]; % <- same elements with 1st row and should be removed

% Process
sort_A = sort(A, 2);
[~, ia] = unique(sort_A, 'stable', 'rows');
A_new = A(ia, :);

% Result
disp(A_new)

     1     2     3
     4     6     5
     1     1     1

[GANESH R]

Its working. Thanks.