In an assignment A(I) = B, the number of elements in B and I must be the same

Dear all,

I am relatively new to MATLAB and trying to work on this code to perform an intensity weighted length measurement on some fluorescence images. My code is here:

%%Load the multi-image
    clear
    [name, path] = uigetfile('.tif');     
    id =strcat(path,name);
    info = imfinfo(id);
    pixsiz=0.16; % to fill in
    num_images = numel(info);
    filename='name';   
    MAT = imread(id); %multi image import:
    Md = double(MAT); 
    [l,w]=size(Md);
    M2 = 1:w;
    M3 = pixsiz*(1:l-2);
    %imshow(Md, []);
%%Process image

% Threshold
    [by,bx]=hist(Md(:),60); %h40,ls=10 
    ft = fittype('gauss1');
    gfit = fit(bx',by',ft);
    coe=coeffvalues(gfit); a1=coe(1); b1=coe(2); c1=coe(3);
    bg2=(b1+0.4*c1); %h40,ls=10
    Mbw1=Md>=bg2; % bw=image>=threshold;
    Mbw2 = Mbw1.*Md;
    %imshow(Mbw2, []);

% Errase noises by labeling connected group
    [LL,num] = bwlabel(Mbw2, 8);
    value = zeros([1,num]);
    for ii=1: num
        temp = LL == ii;
        value(ii) = sum(temp(:));
    end
    [~,index] = max(value);
    Mbw11=LL==index;
    %figure, RGB = label2rgb(LL); imshow(RGB);
    Mbw3 = Mbw11.*Mbw2; %image2 = bw*image;
    %imshow(Mbw3, []);

%%fitting each array of image- 2     
    %st= A(1,1,i)/0.16; %d40 or shorter ls
    b4=ones(1,l);
    Ar=ones(1,l);
    for  k = 1:l
        M1 = Mbw2(k,:);
        %M3 = smooth(M1);
         %plot(M2,M1,'.');
         options = fitoptions('gauss1');
         %options.Upper = [Inf 100 15]; options.Lower = [10 5 0];  options.StartPoint=[8000 st 5]; %d40 or shorter ls
         ft = fittype('gauss1');
         gfit = fit(M2',M1',ft, options);
         coe=coeffvalues(gfit); a2=coe(1); b2=coe(2); c2=coe(3);
         y=a2*exp(-((M2-b2)/c2).^2);
         hold on; plot(M2,y,'r');
         Ar(k) = sqrt(pi)*a2*c2; % Area
         b4(k) = b2; %peak position   
    end

      %%Intensity weighting factor:

      %Lnanoweighting; 
      Ar3=Ar(Ar>0);
      Ar33 = Ar3(1:end-1);
      % For low percentage of coil on DNA
      [hx,hy]=hist(Ar33,30); %A1.histogram of intensity values
      [mx, col]=max(hx);% A2.most probability values of intensity to be a base value 
      Intnew=Ar33/hy(col); % A3.Intensity weighting factor (For low percentage of coil on DNA).
      %  For high percentage of protein_coil on DNA
      %Intnew=Ar33 / min(Ar33);% Intensity weighting factor (For high
      %percentage of protein_coil on DNA).
      % Distance of each pixal (d) 
      AA=Ar>0;
      b32=b4.*AA;
      b33=b32(b32>0);
      l2=length(b33);
      d=ones(1,l2-1);
      B=ones(l2,2); B(:,2)=pixsiz*b33; B(:,1)=pixsiz*(1:l2);
      for p=1:l2-1;
         d(p) =sqrt((B(p+1,1) - B(p,1))^2+(B(p+1,2)-B(p,2))^2);   

      end

      L = sum(d);

      Lnew = L*Intnew';

%%Print results- Lnano, Lnanoweighting
% save the value (area sizes)
    B=ones(2,1); B(2)=L; B(1)=Lnew;
    %outputTXT = regexprep(id, '.dat', ' - Lweighting.txt');
    save([id '.dat'], 'B','-ASCII')

But, when I run this code, I see the following error message:

In an assignment  A(I) = B, the number of elements in B and I must be the
same.

Error in file (line 95)
    B=ones(2,1); B(2)=L; B(1)=Lnew;

Can someone please help me know the source of this problem and a possible solution?

Thanks,

[Thomas Koelen]

Could you show me your what you get when you put

L

and

Lnew

in the command window?

Thomas

Answer 1

[Guillaume]

The error message is fairly clear. You're trying to assign a matrix / or vector (either L, Lnew or both) to a single element of a matrix. This obviously can't work.

Unfortunately, it's fairly difficult to follow your code, not helped by the lack of comments and very non-descriptive variable names. My advice for the future is to use better variable names. For example, RawImage is much clearer than Md, so is ImageHeight, ImageWidth a lot clearer than l and w.

Anyway, the best way for you to solve the problem is to use the debugger. Place a breakpoint in your code and step through it checking the size of variables is what you expect.

Finally, while that will not make a difference for your problem, I would:

• replace strcat by fullfile when building paths.
• replace length by numel for vectors, or size with a explicit dimension for matrices. length on a 2D matrix can either be the number of columns or rows depending on which is greater, so it's never obvious from the code which dimension it returns.
• add a lot more comments to the code.