Why is this code identifying more than one result?

I have written a code which looks through a database and plots certain things at a time. As far as I can tell, having all of my logic checks individually produces the correct locations and number of elements, but when I put them together it tells me that I have more than one result.

para1 = unique([structure.para1].','rows','stable');
for i = 1:length(para1);
  c1 = [structure.para1].'==para1(i);
  para2 = unique([structure(c1).para2].','rows','stable');
  if length(para2)>1;
      hold on
      for j = 1:length(para2);
          c2 = [structure.para2].'==para2(j);
          para3 = unique([structure(c1&c2).para3].','rows','stable');
          for k = 1:length(para3);
              c3 = [structure.para3].'==para3(k);
              set1 = zeros(size(unique(structure(c1&c2&c3).data(:,1)),1),1);
              set1(:) = structure(c1&c2&c3).para3;
              xs = unique(structure(c1&c2&c3).data(:,1));
              for m = 1:length(xs);
                  c4 = structure(c1&c2&c3).data(:,1)==xs(m);
                  qm(m) = max(structure(c1&c2&c3).data(c4,10));
              end
              scatter3(set1,unique(structure(c1&c2&c3).data(:,1)),qm);
              clear set1
          end
      end
      hold off
  end
end

Calling 'structure(c1&c2&c3).data' brings up the specific data block I am looking for, and c4 does correctly logic the elements I am looking for, but when I call

 qm(m) = max(structure(c1&c2&3).data(c4,10));

It returns, 'Expected one output from a curly brace or dot indexing expression, but there were 2 results.' Why is it suddenly looking at multiple data blocks. I suspect it has somehow identified the elements for all para3 values (my actual data has element numbers line up that way), but that makes no sense when c1&c2&c3 does properly identify the single data block I want.

PS if anybody has any better way of reducing this code to not be four for loops I'm open to hearing it.

[dpb]

I can generate both results seemingly with no reason ...

    > [c1 c2 c3]  % show what conditons vectors are
  ans =
    8×3 logical array
     1   1   1
     1   1   0
     1   0   1
     1   0   0
     0   1   1
     0   1   0
     0   0   1
     0   0   0
  >> (c1&c2&3)   % generates the seemingly wrong result
  ans =
    8×1 logical array
     1
     1
     0
     0
     0
     0
     0
     0
  >> c1 & c2   % first piece...looks ok
  ans =
    8×1 logical array
     1
     1
     0
     0
     0
     0
     0
     0
  >> ans&c3     % also ok
  ans =
    8×1 logical array
     1
     0
     0
     0
     0
     0
     0
     0
  >> (c1&(c2&c3))    % group results
  ans =
    8×1 logical array
     1
     0
     0
     0
     0
     0
     0
     0
  >> and(c1,and(c2,c3))  % functional form
  ans =
    8×1 logical array
     1
     0
     0
     0
     0
     0
     0
     0
  >> c1&c2&c3  % back to original -- now it's ok!!??
  ans =
    8×1 logical array
     1
     0
     0
     0
     0
     0
     0
     0
  >> (c1&c2&c3)  % put the parentheses back...no change
  ans =
    8×1 logical array
     1
     0
     0
     0
     0
     0
     0
     0
  >> whos c*
    Name          Size             Bytes  Class      Attributes
    c1            8x1                  8  logical              
    c2            8x1                  8  logical              
    c3            8x1                  8  logical              
    ...  
  >> [c1 c2 c3]  % check haven't inadvertently changed data
  ans =
    8×3 logical array
     1   1   1
     1   1   0
     1   0   1
     1   0   0
     0   1   1
     0   1   0
     0   0   1
     0   0   0
  >> max(structure(c1&c2&)).data(c4,10))
  Expected one output from a curly brace or dot indexing expression, but there were 2 results. 
  >> max(structure(and(c1,and(c2,c3))).data(c4,10))
  ans =
       0
  >>

I can't explain it, but looks like there's something flaky in the input parser with multiple operators in succession without an intermediate store operation.

I think this is a bug; don't see it documented can't write successive operators.

[Walter Roberson]

The code works for me in my tests in R2018a.

Which MATLAB version are you using?

[Stephen23]

max(structure(c1&c2&3).data(c4,10));
                    ^^ not c3

[dpb]

I KNEW that couldn't be right but even while I found one instance I still couldn't get these old eyes to see it in the original...

Thanks, Stephen/Walter...

Answer 1

[Walter Roberson]

(c1&c2&3)

Has &3 as the last component, not &c3

[Bob Thompson]

Wow.

Just wow.

Thanks Walter.