How do I make a (2^m x m) dimensional matrix containing all possible combinations of 1s and 0s?

1 次查看(过去 30 天)
I need to make an n x m matrix with arbitrary m such that n is the number of possible ways to write a m length vector of 1s and 0s (2^m). Example: if m=2, n=4 meaning I have: [1 0], [0 1], [0,0], and [1,1] and can construct the matrix [1 0;0 1;0 0;1 1].
I can not figure out the code to scale this up to higher m in the case where I may have m=3,4,5... and n becomes too large to enter manually. I want to be able to automatically generate the same kind of matrix for any value of m.
This is the Rstudio code someone gave me but I can't translate it:
for(j in 1:m) S <- cbind(S,rep(c(1,0),each=2^(m-j),times=2^(j-1)))
Any help would be much appreciated!
Thank you!

采纳的回答

Matt J
Matt J 2014-2-4
编辑:Matt J 2014-2-4
The lazy way
result = dec2bin(0:2^m-1,m)-'0'

更多回答(1 个)

Matt J
Matt J 2014-2-4
A non-lazy way
[c{m:-1:1}]=ndgrid([0,1]);
result=reshape( cat(m+1,c{:}),[],m)
  4 个评论
Matt J
Matt J 2014-2-5
but if anyone else is interested...
If anyone else is interested, this method is preferable when speed matters. Compare,
m=20;
tic;
[c{m:-1:1}]=ndgrid([0,1]);
result=reshape( cat(m+1,c{:}),[],m) ;
toc
%Elapsed time is 0.090184 seconds.
tic;
result = dec2bin(0:2^m-1,m)-'0';
toc
%Elapsed time is 1.032300 seconds.
Thomas Casey
Thomas Casey 2014-2-5
Yea it worked when I copied and pasted, I must have typed something wrong the first time. That is the fastest way, another person suggested this:
m=3;
mat = nan(2^m,m);
for idx = 1:m
mat(:,idx) = repmat([zeros(2^(m-idx),1);ones(2^(m-idx),1)],2^(idx-1),1);
end;
which is slower but still much much faster than dec2bin.
On my computer, m=3, I got 0.003s (reshape), 0.3s (dec2bin), and 0.0056s (repmat).
Thanks again!

请先登录,再进行评论。

类别

Help CenterFile Exchange 中查找有关 Startup and Shutdown 的更多信息

Community Treasure Hunt

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

Start Hunting!

Translated by