all possible *UNIQUE* permutations of a binary vector

hello all, i am trying to make a griddler/nonogram puzzel solver with matlab.
to do so i need to find all possible UNIQUE permutations of a binary vector. for example :
input Vector: [1 0 1 0]
output matrix:
[1 0 1 0
1 0 0 1
0 1 0 1
0 0 1 1
0 1 1 0
1 1 0 0]
untile now i've been using the following code:
if true
inVec=[1 0 1 0]
outMat=perms(inVec);
outMat=unique(outMat,'rows')
end
This was perfectly fine but for inVec longer then 10 i get an: 'out of memory' error.
is it possible to do this without using perms/unique ? i need this for up to inVec length 20.
Thanks Lampel.

 Accepted Answer

c = [0 1];
cc = c(fullfact([2 2 2 2]));
out = cc(sum(cc,2) == 2,:);
ADD: use Roger Stafford's idea from answer
A = [1 1 0 0];
n = numel(A);
k = sum(A);
c = nchoosek(1:n,k);
m = size(c,1);
out = zeros(m,n);
out(sub2ind([m,n],(1:m)'*[1 1],c)) = 1;

4 Comments

Thanks! Roger Stafford's answer was exactly what I was looking for
A = [1 1 0 0];
n = numel(A);
k = sum(A);
c = nchoosek(1:n,k);
m = size(c,1);
out = zeros(m,n);
% out(sub2ind([m,n],(1:m)'*[1 1],c)) = 1;
out(sub2ind([m,n],(1:m)'*ones(1,k),c
I suggest this minor edit to accomodate any type of A input
I guess instead of
out(sub2ind([m,n],(1:m)'*ones(1,k),c
you meant:
out(sub2ind([m,n],(1:m)'*ones(1,k),c)) = 1;
Also, I have found slightly better performance with
out = zeros(m, n, 'uint8');

Sign in to comment.

More Answers (1)

Jan
Jan on 11 Jun 2013
Edited: Jan on 11 Jun 2013
Using UINT8 instead of DOUBLE will reduce the memroy footprint by a factor of 8.
[EDITED] Bad statistics removed. For 20 elements with 10 enabled bits you get 20!/(10! * (20-10)!) possible solutions, which mean 184'756 * 20 bytes if you use UINT8 values.
Another solution if speed matters: FEX: VChooseK
nElem = 20;
nEnabled = 10;
Index = VChooseK(uint8(1):uint8(nElem), nEnabled);
Result = [under construction], see Andrei's solution

2 Comments

Thanks jan It was very helpfull to know UINT8 reduce the memory print by a factor of 8. permutation of 20 elements leads to huge number. BUT becuse this is a binary vector the unique vectors is not such a large number.
I wanted to try your solution, but VChooseK never finished... :(

Sign in to comment.

Categories

Find more on Graph and Network Algorithms in Help Center and File Exchange

Community Treasure Hunt

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

Start Hunting!