delete same number every two number while keeping the original order

I have one row data, I want the result only leep one number among every two same number. The function"unique" can not do this work. for example: the data i have is as follows:

A = [ 1,1,1,2,2,2,3,3,4,4,4,1,1,1,5,5,2,2,1,1,6]

I want to have the result like this:

A = [1,2,3,4,1,5,2,1,6]. In the result, same number is allowed, but same number every two number is not allowed.

Furthermore, this method is going to be applied to a matrix which have contains number and nan. i.e. B = [1,1,3,3,4,3,3,2,2,nan,nan,nan,nan; 2,3,3,4,4,2,2,nan,nan,nan,nan,nan,nan; 1,1,2,2,1,1,4,4,4,nan,nan,nan,nan]

the result i wish to get is

B = [1,3,4,3,2,nan,nan,nan,nan,nan,nan,nan,nan； 2,3,4,2，nan,nan,nan,nan,nan,nan,nan,nan,nan; 1,2,1,4,nan,nan,nan,nan,nan,nan,nan,nan,nan]

what kind of code can achieve this target? Many thanks!

Answer 1

[Image Analyst]

What about just using the brute force method. It's intuitive and fast:

A = [ 1,1,1,2,2,2,3,3,4,4,4,1,1,1,5,5,2,2,1,1,6]
Aout = A(1); % Initialize output row vector.
for k = 2 : length(A)
  if A(k) ~= Aout(end)
    Aout(end+1) = A(k); % It's a new/different value so add it on.
  end
end
% Print Aout to command window:
Aout

[Andy]

Thank you very much!

Answer 2

[rifat]

a(end+1)=a(end)+1;
x=diff(a);
p=find(x~=0);
a=a(p)

[Andy]

Hi, to apply this code to matrix, i use this code, do u think if there could be any improvement so that the matlab will calculate faster? Thank you very much.

B(:,end+1) = B(end)+1; x =diff(B,1,2); C = []; for temp = 1:size(B,1); p =[]; p=find(x(temp,:)~=0); C(temp,1:length(p)) = B(temp,p); end

C(C ==0)=nan;

Answer 3

[Andrei Bobrov]

x = [true(size(B,1),1), diff(B,1,2)~=0]';
ii = sort(x,'descend');
Bt = B';
z = nan(size(x));
z(ii) = Bt(x);
out = z';