traversing an array of elements

i've the following list of elements as

 block 2=10

 block 5=16

 block 1= 16

 block 7=16

 block 4= 23

 block 6=31

 block 3= 31

i.e i've a total of 7 block numbers

the list is in sorted order...

i want to traverse it and find a match between block values like 'block 5', 'block 1' and 'block 7' match

'block 6' and 'block 3' match.. and print these blocks with matching values

how this can be done?

traversing a single list of array of following type i know like

    function traverse
  clc;
  a=[ 1 2 2 3 3 4 4 4 4]
  len=length(a)-1;
  for i=1:len
      if(a(i)==a(i+1))
          fprintf('\n%d and %d\n',a(i),a(i+1));
      end
  end

but to deal with the above list i dont know help me in doing so....

[Walter Roberson]

Do you mean that you are starting with strings of that form? If so, what form are the strings stored in?

[Raman]

values are stored in array [s,blockNumber]

i want that when this array is traversed and when there occurs a match it will extract the block like here in this case it will extract block 5 and 1 becoz both have same value and after that continues with traversing and where it founds a match extract that block...

in short i want that where there occurs a match extract those blocks i.e list only those blocks

[Walter Roberson]

I do not understand about your "above list of alphabets type" ?

[Raman]

ooooo sorry sir... i have edited it to block... sorry sorry

[Walter Roberson]

You have corrected the length(a) now, as per your recent Question ?

So let a = YourPairsArray(:,1)

where YourPairsArray is the one that has [s,BlockNumber]

[Raman]

sir can u code few lines for me as m not getting it properly

[Walter Roberson]

What is the name of your array that is of the form "[s,blockNumber]" ? Are the values in it stored by columns, so Array(2,1) is the second "s" value?

[Raman]

i've an array of the form

a=[blockNumber(:), S(:)].'

Answer 1

[Jan]

A simple approach:

S = [2, 10;  5, 16;  1, 16;  7, 16;  4, 23;  6, 31;  3, 31];
v = unique(S(:, 2));
for iv = 1:length(v)
  matchIndex = (S(:, 2) == v(iv));
  matchBlock = S(matchIndex, 1)
  fprintf('%d:', v(iv));
  fprintf(' %d', matchBlock);
  fprintf('\n');
end

For larger data sets, histc would be much faster. Search for this command in this forum to find more examples.

[Raman]

i've tried it on my program but it didnt work...

i've divided the input image into 8x8 blocks and after that i've calculated the mean of each and every block and after that i've sorted the mean values and their corresponding blocks value as

[S,blockNumber]=sort(Mean);
fprintf('\nBlock #%d = %d\n', [blockNumber(:), S(:)].' );

the code is as follows:

 a=[blockNumber(:), S(:)].';

 len=length(a)-1;

 v=unique(S(:));

 %disp(v);
 for iv=1:length(v)
 matchIndex=(S(:)==v(iv));
 %disp(matchIndex);
 matchBlock=a(matchIndex,1);
 fprintf('%d:',v(iv));
 fprintf(' %d',matchBlock);
 fprintf('\n');
end

error: index exceeds matrix dimension

can u tell me what is the problem in it? i want the output to be like the output produced by your above code...

[Jan]

Please post the complete error message, especially the line, which causes the error.

[Raman]

sir i have corrected it....

[Jan]

@Jassy: Where did you post the complete error message?