How to delete repeated column in this matrix?

 a= [6.7747   15.1502  126.0000

   67.9227   33.3699   74.0000

   54.9636   40.7709   74.0000

   102.5013  144.4162   44.0000

   67.9227   33.3699   74.0000

   54.9636   40.7709   74.0000];

i should not get the repeated values and i should get like this.

 b=[6.7747   15.1502  126.0000

   67.9227   33.3699   74.0000

   54.9636   40.7709   74.0000

   102.5013  144.4162   44.0000];

[Image Analyst]

There are no repeated columns in that matrix. The example you gave was for a repeated row. Did you mean that instead? Or do you really mean column like you said?

[Matlab111]

sir, check 'a' values

 "[67.9227   33.3699   74.0000
   54.9636   40.7709   74.0000]"

this two is repeated.

[Niels]

I believe he refers to the fact that a matrix consist of rows and columns and that your example is, in contrast to what you state, referring to duplicate rows instead of columns ;)

[Matlab111]

ya that's right it's not a matrix it's a value that's consist three variables like this

 'x'          y         z
[67.9227   33.3699   74.0000];

but i'm considering only rows not columns in above my example.

Answer 1

[Niels]

b = unique(a,'rows','stable');

b =
                  6.7747                   15.1502                       126
                 67.9227                   33.3699                        74
                 54.9636                   40.7709                        74
                102.5013                  144.4162                        44

That should do it.

[Matlab111]

This is not working...

[Cedric]

Why? Be more explicit.

[Niels]

From your response I assume that you have a matlab version older than R2012a.

In that case, you can do the following:

[~,x,~] = unique(a,'rows','first');
b = a(sort(x),:);

And if you do not care about the order of your rows, simply doing unique(a,'rows') suffices.

[Matlab111]

still i'm not getting correct answer. k just see my code i attached below, just run my code, you get that repeated matrix that's the matrix what i discussed above. so in this i should not get repeated rows.

[Niels]

I checked your code and you seem to be doing something wrong in there.

If you actually have a matrix where you take the unique rows from, it will work. But instead your S_nodes variable is a vector which is overwritten in every iteration in the for loop @ line 308-325.

It seems like the numbers you print to the screen are the rows you obtain. What you should do is:

1. Store these rows in a large matrix for every iteration.
2. After all iterations are done, perform the unique operation and eliminate the duplicates.

[Matlab111]

ya i understand what your saying and i tried that i'm not getting and actually i don't know how to store that rows.

[Niels]

If you know beforehand how many rows you will obtain, you could do something like this:

allRows = zeros(N,3); % put this before line 308 where your loop starts. N is the number of rows you will expect.

Then replace your line b = S_nodes(sort(x),:); with

allRows(i,:) = S_nodes;

And finally place your uniqueness check after line 325 (where the loop ends):

b = unique(allRows,'rows','stable');
disp(b);

or

[~,x,~] = unique(S_nodes,'rows','first');
b = S_nodes(sort(x),:);
disp(b);

If you don't know how many rows you will obtain, you can do it in a slightly more dirty way:

allRows = []; % put this before line 308 where your loop starts. N is the number of rows you will expect.

And in your loop:

allRows(end+1,:) = S_nodes;

The rest can be done in the same fashion as above.

This should be sufficient to get you started.

[Matlab111]

ya Thank you i'm getting...