How to use ISMEMBER when not all rows match

I'm trying to dynamically update and compare values corresponding to their respective ID. This problem is easy if all ID's are same each time (use horzcat). The problem is that not all ID's are represented each time. For these, I want the value corresponding to the ID to be 0. See simple example below:

A=[1 2 3 4 5 6; 1 1 1 1 1 1]'
B=[2 3 4;2 2 2]'

where A(:,1) and B(:,1) refer to the ID; so A and B share 2 3 4 in common and A(:,2) and B(:,2) are values in two different time periods, i.e., value of ID=2 is 1 in t=1 and 2 in t=2. But elements 1, 5, 6 didn't occur in time period 2.

I want the result:

compare=[1 2 3 4 5 6;1 1 1 1 1 1;0 2 2 2 0 0]'

So it fills in zeros if the ID is not represented in B. If you use:

[~,idx]=ismember(A,B,'rows')

you get:

idx =

             0
          1.00
          2.00
          3.00
             0
             0

But rather than the index, I want it to fill in the values of B(:,2) that match in their respective row.

Answer 1

[Titus Edelhofer]

Hi Sophia,

what about this?

compare = [A zeros(size(A,1),1)]
[ismemb,idx] = ismember(A(:,1), B(:,1))
compare(ismemb,3) = B(idx(ismemb))

Does this help?

Titus

[Sophia]

Thank you, this is very helpful!