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

My code is like

I=zeros(1,5);
    for i=1:4
    [ma,I(i)]=max(areas);
end

Error: "In an assignment A(I) = B, the number of elements in B and I must be the same." [ma,I(i)]=max(areas); the second argument y in [x,y]=max(a) is a single element and I am trying to send single element only right?

Answer 1

[Walter Roberson]

max() of a 2D array results in a vector, not a scalar. The indices of that would not fit in the single location I(i)

You should be asking yourself why you are doing the same thing in every iteration of the loop. max(areas) is not going to change in the loop.

Why are you allocating 5 locations but only looping to 4?

[pranith kumar]

Thanks for the reply.

The original code is different and very long. max(areas) is going to change in each loop as per that code. I just gave a small part to show where my error mainly is.

sorry, I should have given the question in better format.

my areas is a row vector.

    I=zeros(1,4);

    for i=1:4

    [ma,I(i)]=max(areas);
     areas(I(i))=min(areas);
     end

so I am trying to get the Indices of first four Max areas in to I vector. now please give me a code for that please.

[Walter Roberson]

Your area variable is becoming a 2D array at some point.

Just before that section of code, add

   assert(isscalar(areas), 'areas is unexpectedly array %s', str2num(size(areas)));

Answer 2

[Guillaume]

If areas is anything but a vector or scalar, then the return values of max are not scalar. See:

areas = magic(3)
[ma, l] = max(areas)

If you want the max of the whole matrix:

[ma, l(i)] = max(areas(:));

Note that in that case l(i) is the linear index of the max location.

[pranith kumar]

sorry, I should have given the question in better format. my areas is a ROW VECTOR .

I=zeros(1,4);

    for i=1:4

    [ma,I(i)]=max(areas);
     areas(I(i))=min(areas);
end

so I am trying to get the Indices of first four Max areas in to I vector. now please give me a code for that please.