Unable to perform assignment because the indices on the left side are not compatible with the size of the right side.

for i=1:(Num_Ele+Num_Res);
   for j=1:(Num_Ele+Num_Res);
      for k=1:length(conn);
         ind1=conn(k,1);
         ind2=conn(k,2);
         F(i)=xy(ind2,:)-xy(ind1,:);
         A(i,j)= acos(dot(F(k,:),[1,0])/norm(F(k,:)));
      end
   end
end

Answer 1

[Adam Danz]

You're trying to put x items in y spaces. For example, in the lines below I'm trying to put 3 numbers in 1 space.

x = [10,11,12,13,14]; 
x(1) = [7,8,9];  % Error!

Here's another example where I try to put 3 items in two spaces.

x = {'d' 'e' 'f' 'g' 'h'};
x(1:2) = {'a','b','c'}; % Error!

One more example with a matrix. Here I try to replace the first row [1,2,3] with [1,2,3,4]

x = [1 2 3; 
     4 5 6; 
     7 8 9]; 
x(1,:) = [1 2 3 4]; % Error!

These are indexing errors.

In your case, if F and A are vectors or matrices then you can only store one value (a scalar) in F(i) and one value in A(i,j). Without knowing what your variable are, I can only guess that this line below produces a vector.

xy(ind2,:)-xy(ind1,:);

So you might be trying to store >1 value in F(i). The same is probably true to A(i,j).

If F and A are cell arrays, then you can store a vector or matrix (and more) into a single element F{i}

So, you need to look at the values you are attempting to store in F(i) and A(i,j) to determine if they are scalars or not.

Answer 2

[Luca Parziale]

I have the same error message :

Unable to perform assignment because the indices on the left side are not compatible with the size of the right side.

But for this code :

for i = 1:numImages
    img = readimage(trainingSet,i);
    
    img = im2gray(img);
    
    % Apply pre-processing steps
    img = imbinarize(img);
    
    trainingFeatures(i, :) = extractHOGFeatures(img,'CellSize',cellSize);  % cellSize = [4 4];
end

I can make 1030 itéreations and then I get the error message on this line :

 trainingFeatures(i, :) = extractHOGFeatures(img,'CellSize',cellSize);  % cellSize = [4 4];

The code is this one (from de matlab documentation) :

https://ch.mathworks.com/help/vision/ug/digit-classification-using-hog-features.html

It's exactly the same but with my own dataset of 30x30 images. And I have 5832 images 30x30 in my Dataset.

When I run the code from the link abow (matlab documentation code) it works well but with my dataset it doesn't work and I don't know why ? Could someone give an hint or something to help me please ?

[Walter Roberson]

I predict that one of the images is not exactly the same size as the others. You should

dbstop if error

and check size(img)

[Luca Parziale]

Thank you. Indeed, I hade one image over 5832 that was 30x26 instead of 30x30 !