How can I create a set of N diagnal matrices range from diag(1, 1, ... , 1) to diag(-1, -1, ... , -1) for testing purposes?

1 view (last 30 days)
I am needing to create 16 4x4 diagnal matrices consisting of all of the combinations of "1" and "-1" and have them be able to be called for use in a loop. Besides creating each individual matrix, how can I go about accomplishing this?
for instance:
Z1 = diag(1,1,1,1);
.
Z8 = diag(1,-1,-1,1);
.
Z16 = diag(-1,-1,-1,-1);
Thank you for any guidance.
  3 Comments
Zac Minson
Zac Minson on 20 Sep 2021
I ended up doing this, but was wondering if there is a more general method in the case of 2^N matrices.
Stephen23
Stephen23 on 21 Sep 2021
@Zac Minson: using a cell array is a very general method, which works for any number of matrices.

Sign in to comment.

Accepted Answer

Matt J
Matt J on 20 Sep 2021
Edited: Matt J on 21 Sep 2021
Do you really need all Z{i} at the same time?
[C{1:N}]=ndgrid([-1,1]);
C=reshape( cat(N+1,C{:}) ,[],N);
Z=zeros(N^2,2^N); %EDITED - removed loops
Z(1:N+1:N^2,:)=C;
Z=reshape(Z, N,N,[])
  3 Comments
Matt J
Matt J on 20 Sep 2021
Here's a loop-free version:
N = 4; % Set this value
[C{1:N}]=ndgrid([-1,1]);
C=reshape( cat(N+1,C{:}) ,[],N).';
Z=zeros(N^2,2^N);
Z(1:N+1:N^2,:)=C;
Z=reshape(Z, N,N,[])
Z =
Z(:,:,1) = -1 0 0 0 0 -1 0 0 0 0 -1 0 0 0 0 -1 Z(:,:,2) = 1 0 0 0 0 -1 0 0 0 0 -1 0 0 0 0 -1 Z(:,:,3) = -1 0 0 0 0 1 0 0 0 0 -1 0 0 0 0 -1 Z(:,:,4) = 1 0 0 0 0 1 0 0 0 0 -1 0 0 0 0 -1 Z(:,:,5) = -1 0 0 0 0 -1 0 0 0 0 1 0 0 0 0 -1 Z(:,:,6) = 1 0 0 0 0 -1 0 0 0 0 1 0 0 0 0 -1 Z(:,:,7) = -1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 -1 Z(:,:,8) = 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 -1 Z(:,:,9) = -1 0 0 0 0 -1 0 0 0 0 -1 0 0 0 0 1 Z(:,:,10) = 1 0 0 0 0 -1 0 0 0 0 -1 0 0 0 0 1 Z(:,:,11) = -1 0 0 0 0 1 0 0 0 0 -1 0 0 0 0 1 Z(:,:,12) = 1 0 0 0 0 1 0 0 0 0 -1 0 0 0 0 1 Z(:,:,13) = -1 0 0 0 0 -1 0 0 0 0 1 0 0 0 0 1 Z(:,:,14) = 1 0 0 0 0 -1 0 0 0 0 1 0 0 0 0 1 Z(:,:,15) = -1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1 Z(:,:,16) = 1 0 0 0 0 1 0 0 0 0 1 0 0 0 0 1
Zac Minson
Zac Minson on 21 Sep 2021
Thank you so much. I was really wracking my brain over how to generalize this problem. I appreciate your time.

Sign in to comment.

More Answers (2)

Fangjun Jiang
Fangjun Jiang on 20 Sep 2021
a=(dec2bin(15:-1:0)-48)*2-1;
z=zeros(4,4,16);
for k=1:16
z(:,:,k)=diag(a(k,:));
end
  1 Comment
the cyclist
the cyclist on 20 Sep 2021
Here is a generalization of this solution:
N = 4; % Set this value
a=(dec2bin((2^N-1):-1:0)-48)*2-1;
z=zeros(N,N,2^N);
for k=1:2^N
z(:,:,k)=diag(a(k,:));
end

Sign in to comment.


Jan
Jan on 20 Sep 2021
Edited: Jan on 21 Sep 2021
z = repmat(eye(4), 1, 1, 16);
z(z==1) = 1 - rem(floor((0:15) ./ [1; 2; 4; 8]), 2) * 2; % Inlined DEC2BIN
Generalized:
n = 3;
z = repmat(eye(n), 1, 1, 2^n);
z(z==1) = 1 - rem(floor((0:2^n-1) ./ pow2(0:n-1).'), 2) * 2;
Now z(:,:,i) is the i.th matrix. This is more efficient than storing them in a cell. But if you want this:
zC = num2cell(z, [1,2]);

Products


Release

R2021a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!