Question

Suppose now I have a matrix

S = [1 1 1 2 2 2;
     1 1 1 2 2 2;
     2 2 2 2 1 1;
     2 2 2 2 1 1;
     2 2 2 2 1 1]

And another matrix

A = [1 2;
     2 4]

The first row in A is the unique indices of S, and the second row contains the values that the values in the first row will be replaced. That is, all “1”s in S will be replaced by 2, and all “2”s will be replaced by 4. Finally I’ll get a matrix

SS = [2 2 2 4 4 4;
      2 2 2 4 4 4;
      4 4 4 4 2 2;
      4 4 4 4 2 2;
      4 4 4 4 2 2]

Right now what I’m doing is:

SS = zeros(size(S));
for i = 1:size(A,2)
    SS(S==index(A(1, i)) = A(2,i);
end

Now, I have a pretty big matrix, and using a for loop is a little bit slow. Is there a faster way to do that?

Answer 1

Use the second output of ismember to give you indices of the values in row 1 of A. Use these indices to directly create matrix SS.

Example (changed initial values for clarity):

S = [5 5 5 3 3 3;
5 5 5 3 3 3;
3 3 3 3 5 5;
3 3 3 3 5 5;
3 3 3 3 5 5];
A = [5 3;
2 4];

>> [~, Locb] = ismember(S,A(1,:))
Locb =

     1     1     1     2     2     2
     1     1     1     2     2     2
     2     2     2     2     1     1
     2     2     2     2     1     1
     2     2     2     2     1     1

>> SS = reshape(A(2,Locb),size(S))
SS =

     2     2     2     4     4     4
     2     2     2     4     4     4
     4     4     4     4     2     2
     4     4     4     4     2     2
     4     4     4     4     2     2