sum over matrix columns

Hi all, I have the following problem:

I defined the matrix A: A=[1 2 1; 4 5 1] then with sum(A) I got the desired sums calculated: 5, 7, 2. So far so fine. After that I executed the following command: sum((bsxfun(@minus,x,permute(X,[1,3,2]))/h).^2), where x=[1 2 3 6 7; 4 5 2 1 6] and X=[1 2 1; 4 5 3], h=0.1 and got the following error message: "Subscript indices must either be real positive integers or logicals." This error occures now everytime I want to execute A=[1 2 3; 4 5 6; 1 0 1] and sum(A). What did I wrong?

Do you maybe know what might be a reason for such behaviour? Thnak you Best Regards Xanka

[David Young]

Difficult to know - I don't get the error with your code:

x=[1 2 3 6 7; 4 5 2 1 6];
X=[1 2 1; 4 5 3];
h = 0.1;
sum((bsxfun(@minus,x,permute(X,[1,3,2]))/h).^2)
ans(:,:,1) =
         0         200         800        3400        4000
ans(:,:,2) =
       200           0        1000        3200        2600
ans(:,:,3) =
       100         500         500        2900        4500

How are you executing your code? Is it in an m-file or are you pasting it into the command window? Are you sure that you are only executing the correct lines of code?

For trying ideas out, I recommend using cell mode in the editor.

Answer 1

[xanka bocem]

Thank you all.

I found the mistake. I used sum as variable and that´s why I got then the error when I tried to use sum as a function. Best Regards Xanka

Answer 2

[Azzi Abdelmalek]

Check if permute was not used as a variable in your code. Type

permute