The sum of stacked matrices

I have a 3-d matrix and I want to find the sum so that every element in the sum matrix is the sum of all the elements in that corresponding i,j location. I know that sounds confusing probably so for example:

>> A=zeros(3,3,3);
>> A(:,:,1)=[1 1 1; 1 0 1; 0 0 0];
>> A(:,:,2)=[1 1 1; 0 0 1; 0 1 0];
>> A(:,:,3)=[1 0 1; 0 1 0; 0 1 0];

Would give me a sum matrix that would be:

 [ 3 2 3
   1 1 2
   0 2 0 ]

I realize I could reshape the matrices into a vector, sum them, and then reshape them back into the matrix, and that I could use for loops, but both of those seem over complicated and I would be doing this with a pretty large amount of data so speed is ideal :)

Thanks in advance!

-Shaun

Answer 1

[Roger Stafford]

sum(A,3)

Roger Stafford

[Shaun VanWeelden]

Well how bout that, that was sure easy haha

Question though, how does that 3 work? I was looking in the sum documentation and found that it sums along the second argument as the dimension, but I guess I don't understand what this means. Could you help explain it?

And by a similiar note, I tried finding the entire sum of a 2d matrix by doing sum(a2dMat,2) to see if it would give me just one value as the sum of all elements, but that did not work so now I'm a little confused as to how the second arg works in Sum

[Roger Stafford]

No, it sums along the third dimension, which is what you asked for. That is what the '3' refers to.

To get the entire sum of a two-dimensional array you need to either use two calls to 'sum' or convert the array to one-dimension and sum that:

 sum(sum(A))

or

 sum(A(:))

Roger Stafford